GRINDING METHOD AND GRINDING MEDIUM

An improved grinding method for manufacturing microfibrillated cellulose.

Skip to: Description  ·  Claims  · Patent History  ·  Patent History
Description
TECHNICAL FIELD

The present invention is directed to an improved grinding method for manufacturing microfibrillated cellulose.

BACKGROUND OF THE INVENTION

Grinding methods and compositions for the production of compositions comprising microfibrillated cellulose are described in WO2010/131016, WO2015/173376, and WO2018193314. Paper products comprising such microfibrillated cellulose have been shown to exhibit excellent paper properties, such as paper strength.

Despite the benefits seen in WO2010/131016, there is ongoing need to improve the economics of producing microfibrillated cellulose on an industrial scale, and to develop new grinding processes for producing microfibrillated cellulose. It is also desirable to be able to enhance one or more mechanical properties of microfibrillated cellulose, for example, its tensile index and/or tensile strength. It is also desirable to further improve the manufacture of microfibrillated cellulose utilizing a grinding process that minimizes wearing of the grinding media thereby reducing the level of worn grinding media in the final microfibrillated cellulose product.

SUMMARY OF THE INVENTION

According to a first aspect, the present invention is directed to a method for preparing microfibrillated cellulose having at least one of increased tensile properties and reduced media wear, said method comprising: providing a suspension comprising pulp fibres in a liquid medium, wherein the fibre solids content of the suspension is about 0.3 wt % to about 4 wt %; microfibrillating the pulp fibres by grinding the suspension in the presence of a grinding medium in one or more wet stirred media mills, wherein: the grinding medium comprises about 0.5 mm to about 12 mm particles, the grinding medium has an arithmetic surface roughness from about 0.02 μm to about 2 μm, the media volume concentration of the grinding media is about 30% to about 65% of the total volume of the charge of the one or more wet stirred media mills, the grinding is performed with an energy input of less than about 6,000 kWh/t, and grinding the suspension results in a worn media content of less than about 20% by dry mass; and recovering the microfibrillated cellulose from the one or more wet stirred media mills.

In some embodiments of the first aspect, the liquid medium is aqueous.

In some embodiments of the first aspect, the liquid medium is a hydro alcoholic mixture.

In some embodiments of the first aspect, the pulp fibres are hardwood pulp fibres.

In some embodiments of the first aspect, the hardwood pulp fibres are selected from the group consisting of acacia, albizia, alder, anthocephalus, ash, beech, birch, catalpa, cherry, cottonwood, elm, eucalyptus, gum, gmelina, hickory, locust, magnolia, maple, oak, poplar, sassafras, sycamore, walnut, and mixtures thereof.

In some embodiments of the first aspect, the hardwood pulp fibres are eucalyptus.

In some embodiments of the first aspect, the hardwood pulp fibres are acacia.

In some embodiments of the first aspect, the hardwood pulp fibres are birch.

In some embodiments of the first aspect, the hardwood pulp fibres are mixed North American hardwoods.

In some embodiments of the first aspect, the hardwood pulp fibres are mixed Northern European hardwoods.

In some embodiments of the first aspect, the hardwood pulp fibres are mixed Southeast Asian hardwoods.

In some embodiments of the first aspect, the hardwood pulp fibres are a tropical hardwood pulp fibre.

In some embodiments of the first aspect, the pulp fibres are softwood pulp fibres.

In some embodiments of the first aspect, the softwood pulp fibres are selected from the group consisting of Northern Softwood Kraft (NSK) pulp fibres, Southern Softwood Kraft (SSK) pulp fibres, Tropical Softwood Kraft (TSK) pulp fibres, Northern Bleached Softwood Kraft (NBSK) pulp fibres, southern pines, red cedar, hemlock, black spruce, and mixtures thereof.

In some embodiments of the first aspect, the softwood pulp fibres are Northern Softwood Kraft (NSK) pulp fibres.

In some embodiments of the first aspect, the softwood pulp fibres are Southern Softwood Kraft (SSK) pulp fibres.

In some embodiments of the first aspect, the softwood pulp fibres are Tropical Softwood Kraft (TSK) pulp fibres.

In some embodiments of the first aspect, the softwood pulp fibres are mixed softwood pulp fibres.

In some embodiments of the first aspect, the softwood pulp fibres are Northern Bleached Softwood Kraft (NBSK) pulp fibres.

In some embodiments of the first aspect, the pulp fibres are recycled pulp fibres.

In some embodiments of the first aspect, the pulp fibres are non-wood cellulosic fibres.

In some embodiments of the first aspect, the non-wood cellulosic fibres are selected from the group consisting of bagasse, abaca, kenaf, sisal, cotton, hemp, flax, miscanthus, sorghum, jute, bamboo, and mixtures thereof.

In some embodiments of the first aspect, the pulp fibres are a combination of two or more of hardwood pulp fibres, softwood pulp fibres, recycled pulp fibres, and non-wood cellulosic pulp fibres.

In some embodiments of the first aspect, the fibre solids content of the suspension is about 0.3 wt. %, about 0.4 wt. %, about 0.5 wt. %, about 0.6 wt. %, about 0.7 wt. %, about 0.8 wt. %, about 0.9 wt. %, about 1 wt. %, about 1.1 wt. %, about 1.2 wt. %, about 1.3 wt. %, about 1.4 wt. %, about 1.5 wt. %, about 1.6 wt. %, about 1.7 wt. %, about 1.8 wt. %, about 1.9 wt. %, about 2 wt. %, about 2.1 wt. %, about 2.2 wt. %, about 2.3 wt. %, about 2.4 wt. %, about 2.5 wt. %, about 2.6 wt. %, about 2.7 wt. %, about 2.8 wt. %, about 2.9 wt. %, about 3 wt. %, about 3.1 wt. %, about 3.2 wt. %, about 3.3 wt. %, about 3.4 wt. %, about 3.5 wt. %, about 3.6 wt. %, about 3.7 wt. %, about 3.8 wt. %, about 3.9 wt. %, about or 4 wt. %.

In some embodiments of the first aspect, the fibre solids content of the suspension is about 0.5 wt. % to about 2 wt. %.

In some embodiments of the first aspect, the one or more wet stirred media mills comprises one or more vertical screened grinders.

In some embodiments of the first aspect, the one or more wet stirred media mills comprises one or more horizontal milling apparatuses

In some embodiments of the first aspect, the one or more wet stirred media mills comprises one or more stirred media detritors.

In some embodiments of the first aspect, the one or more wet stirred media mills is two or more wet stirred media mills connected in series, wherein the two or more wet stirred media mills are selected from the group consisting of a screened grinder, stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is three or more wet stirred media mills connected in series, wherein the three or more wet stirred media mills are selected from the group consisting of a screened grinder, stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is four or more wet stirred media mills connected in series, wherein the four or more wet stirred media mills are selected from the group consisting of a screened grinder, a stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is five or more wet stirred media mills connected in series, wherein the five or more wet stirred media mills are selected from the group consisting of a screened grinder, a stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is six or more wet stirred media mills connected in series, wherein the six or more wet stirred media mills are selected from the group consisting of a screened grinder, a stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is from two to ten wet stirred media mills connected in series, wherein the two to ten wet stirred media mills are selected from the group consisting of a screened grinder, a stirred media detritor, a horizontal milling apparatus, and combinations thereof.

In some embodiments of the first aspect, the one or more wet stirred media mills is first and second wet stirred media mills, and the first wet stirred media mill includes grinding medium having a particle size greater than a particle size of grinding medium included in the second wet stirred media mill.

In some embodiments of the first aspect, the one or more wet stirred media mills is first and second wet stirred media mills, and the first wet stirred media mill includes grinding medium having a particle size substantially similar to a particle size of grinding medium included in the second wet stirred media mill.

In some embodiments of the first aspect, the one or more wet stirred media mills is two or more wet stirred media mills connected in series, and an output of a first of the two or more wet stirred media mills has a fibre breakage factor of 0.2 mm−1-3 mm−1 as calculated using data measured by a fibre analyzer (e.g., a Valmet FS5 fibre image analyzer).

In some embodiments of the first aspect, the fibre breakage factor is about 0.2 mm−1, about 0.3 mm−1, about 0.4 mm−1, about 0.5 mm−1, about 0.6 mm−1, about 0.7 mm−1, about 0.8 mm−1, about 0.9 mm−1, about 1.0 mm−1, about 1.1 mm−1, about 1.2 mm−1, about 1.3 mm−1 about 1.4 mm−1, about 1.5 mm−1, about 1.6 mm−1, about 1.7 mm−1, about 1.8 mm−1, about 1.9 mm−1, about 2.0 mm−1, about 2.1 mm−1, about 2.2 mm−1, about 2.3 mm−1, about 2.4 mm−1, about 2.5 mm−1, about 2.6 mm−1, about 2.7 mm−1, about 2.8 mm−1, about 2.9 mm−1, or about 3.0 mm−1.

In some embodiments of the first aspect, the grinding medium comprises about 0.5 mm, about 0.6 mm, about 0.7 mm, about 0.8 mm, about 0.9 mm, about 1 mm, about 1.1 mm, about 1.2 mm, about 1.3 mm, about 1.4 mm, about 1.5 mm, about 1.6 mm, about 1.7 mm, about 1.8 mm, about 1.9 mm, about 2 mm, about 2.1 mm, about 2.2 mm, about 2.3 mm, about 2.4 mm, about 2.5 mm, about 2.6 mm, about 2.7 mm, about 2.8 mm, about 2.9 mm, about 3 mm, about 3.1 mm, about 3.2 mm, about 3.3 mm, about 3.4 mm, about 3.5 mm, about 3.6 mm, about 3.7 mm, about 3.8 mm, about 3.9 mm, about 4 mm, about 4.1 mm, about 4.2 mm, about 4.3 mm, about 4.4 mm, about 4.5 mm, about 4.6 mm, about 4.7 mm, about 4.8 mm, about 4.9 mm, about 5 mm, about 5.1 mm, about 5.2 mm, about 5.3 mm, about 5.4 mm, about 5.5 mm, about 5.6 mm, about 5.7 mm, about 5.8 mm, about 5.9 mm, about 6 mm, about 6.1 mm, about 6.2 mm, about 6.3 mm, about 6.4 mm, about 6.5 mm, about 6.6 mm, about 6.7 mm, about 6.8 mm, about 6.9 mm, about 7 mm, about 7.1 mm, about 7.2 mm, about 7.3 mm, about 7.4 mm, about 7.5 mm, about 7.6 mm, about 7.7 mm, about 7.8 mm, about 7.9 mm, about 8 mm, about 8.1 mm, about 8.2 mm, about 8.3 mm, about 8.4 mm, about 8.5 mm, about 8.6 mm, about 8.7 mm, about 8.8 mm, about 8.9 mm, about 9 mm, about 9.1 mm, about 9.2 mm, about 9.3 mm, about 9.4 mm, about 9.5 mm, about 9.6 mm, about 9.7 mm, about 9.8 mm, about 9.9 mm, about 10 mm, about 10.1 mm, about 10.2 mm, about 10.3 mm, about 10.4 mm, about 10.5 mm, about 10.6 mm, about 10.7 mm, about 10.8 mm, about 10.9 mm, about 11 mm, about 11.1 mm, about 11.2 mm, about 11.3 mm, about 11.4 mm, about 11.5 mm, about 11.6 mm, about 11.7 mm, about 11.8 mm, about 11.9 mm, or about 12 mm particles.

In some embodiments of the first aspect, the grinding medium comprises about 0.5 mm to about 8 mm particles.

In some embodiments of the first aspect, the grinding medium comprises about 1 mm to about 5 mm particles.

In some embodiments of the first aspect, the grinding medium comprises about 1 mm to about 3 mm particles.

In some embodiments of the first aspect, the grinding medium comprises mullite, alumina, silicate, zirconia, glass, steatite, or a combination thereof.

In some embodiments of the first aspect, the grinding medium comprises mullite.

In some embodiments of the first aspect, the grinding medium comprises mullite and alumina.

In some embodiments of the first aspect, the grinding medium has an arithmetic surface roughness of about 0.02 μm, about 0.03 μm, about 0.04 μm, about 0.05 μm, about 0.06 μm, about 0.07 μm, about 0.08 μm, about 0.09 μm, about 0.1 μm, about 0.2 μm, about 0.3 μm, about 0.4 μm, about 0.5 μm, about 0.6 μm, about 0.7 μm, about 0.8 μm, about 0.9 μm, about 1 μm, about 1.1 μm, about 1.2 μm, about 1.3 μm, about 1.4 μm, about 1.5 μm, about 1.6 μm, about 1.7 μm, about 1.8 μm, about 1.9 μm, or about 2 μm.

In some embodiments of the first aspect, the grinding medium has an arithmetic surface roughness of about 0.1 μm to about 1 μm.

In some embodiments of the first aspect, the grinding medium has a density of about 2 g/cm3 to about 8 g/cm3.

In some embodiments of the first aspect, the grinding is performed with an energy input of about 500 kWhH/t, about 600 kWhH/t, about 700 kWhH/t, about 800 kWhH/t, about 900 kWhH/t, about 1,000 kWhH/t, about 1,100 kWhH/t, about 1,200 kWhH/t, about 1,300 kWhH/t, about 1,400 kWhH/t, about 1,500 kWhH/t, about 1,600 kWhH/t, about 1,700 kWhH/t, about 1,800 kWhH/t, about 1,900 kWhH/t, about 2,000 kWhH/t, about 2,100 kWhH/t, about 2,200 kWhH/t, about 2,300 kWhH/t, about 2,400 kWhH/t, about 2,500 kWhH/t, about 2,600 kWhH/t, about 2,700 kWhH/t, about 2,800 kWhH/t, about 2,900 kWhH/t, about 3,000 kWhH/t, about 3,100 kWhH/t, about 3,200 kWhH/t, about 3,300 kWhH/t, about 3,400 kWhH/t, about 3,500 kWhH/t, about 3,600 kWhH/t, about 3,700 kWhH/t, about 3,800 kWhH/t, about 3,900 kWhH/t, about 4,000 kWhH/t, about 4,100 kWhH/t, about 4,200 kWhH/t, about 4,300 kWhH/t, about 4,400 kWhH/t, about 4,500 kWhH/t, about 4,600 kWhH/t, about 4,700 kWhH/t, about 4,800 kWhH/t, about 4,900 kWhH/t, about 5,000 kWhH/t, about 5,100 kWhH/t, about 5,200 kWhH/t, about 5,300 kWhH/t, about 5,400 kWhH/t, about 5,500 kWhH/t, about 5,600 kWhH/t, about 5,700 kWhH/t, about 5,800 kWhH/t, about 5,900 kWhH/t, or about 6,000 kWhH/t.

In some embodiments of the first aspect, the grinding is performed with an energy input of less than about 5,000 kWh/t.

In some embodiments of the first aspect, the media concentration of the grinding medium is about 30%, about 35%, about 40%, about 45%, about 50%, about 55%, about 60%, or about 65% of the total volume of the charge of the one or more wet stirred media mills.

In some embodiments of the first aspect, the media volume concentration of the grinding medium is about 35% to about 55% of the total volume of the charge of the one or more wet stirred media mills.

In some embodiments of the first aspect, the worn media content is about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 1%, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, about 10%, 11%, about 12%, about 13%, about 14%, about 15%, about 16%, about 17%, about 18%, about 19%, or about 20%.

In some embodiments of the first aspect, the worn media content is less than about 10%.

In some embodiments of the first aspect, the microfibrillated cellulose has a FLT value of greater than 7.

In some embodiments of the first aspect, the microfibrillated cellulose has a FLT value greater than abut 8, greater than bout 9, greater than 10, greater than 11, greater than 12, greater than 13, greater than 14, or greater than 15.

In some embodiments of the first aspect, the microfibrillated cellulose has a FLT value of about 8, about 9, about 10, about 11, about 12, about 13, about 14, or about 15.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1a is a plot of Lc (I) vs energy input at 100 POP.

FIG. 1b is a plot of FLT vs energy input at 100 POP.

FIG. 2 is a plot of pilot-scale 50 POP batch grinds with NBSK, comparing 4 different media sizes.

FIG. 3a is a plot of Lc (I) vs. specific energy input for a lab grind of eucalyptus at 100 POP.

FIG. 3b is a plot of FLT vs. specific energy input for a lab grind of eucalyptus at 100 POP.

FIG. 4 is a comparison of 3 mm and 1.7 mm media continuous grinds for Eucalyptus at pilot scale.

FIG. 5 provides data for a full scale continuous grind of eucalyptus at 100 POP

FIG. 6 provides data for a full scale continuous grind of eucalyptus at 100 POP

FIG. 7 provides data from a pilot plant continuous grinding of birch at 100 POP.

FIGS. 8a and 8b show data from batch lab grinds with 3 mm and 1.7 mm media for acacia fibres. FIG. 8a shows Lc (I) vs. energy input. FIG. 8b shows FLT vs. energy input.

FIG. 9 shows pilot-scale batch acacia grinds at 50 POP.

FIG. 10 is a plot of MFC tensile index for various fibres ground with 1 and 2.9 mm media.

FIG. 11 is a plot of the ratio of MFC tensile index between fine and coarse media series versus fibre zero-span tensile index.

FIG. 12 is a plot of Lc (I)-based operating index versus feed fibre length for 2.9 and 1 mm media grinds.

FIG. 13 is a plot of MFC Lc (I) values for various fibre species ground with 1 and 2.9 mm media.

FIGS. 14a through 14h are differential interference contrast microscopy images of MFC produced by 1 mm and 2.9 mm media grinds of various fibre species.

FIG. 15a is a plot of Lc (I) vs. energy input for batch lab grinds with NBSK comparing 3 mm and 1.7 mm media.

FIG. 15b is a plot of FLT vs. energy input for batch lab grinds with NBSK comparing 3 mm and 1.7 mm media

FIG. 16 is a plot of full-scale continuous grinds with several media sizes comparing 100 POP to 50 and 20 POP.

FIG. 17 is a plot of Lc (I) vs. energy input for full-scale continuous grinds with NBSK.

FIG. 18 is a plot of Lc (I) vs. energy input at full scale with eucalyptus in continuous mode, comparing 100 POP and 50 POP grinds.

FIG. 19 is a plot of pilot-scale continuous grinds with acacia fibres, comparing 100, 50, and 25 POP grinds with 1.7 mm media.

FIG. 20 is a plot of 4-stage cascade grinding at pilot scale in continuous mode with NBSK at 100 POP, showing the FLT-energy relationship at each stage. The square symbol is the single stage 3 mm media 50 POP ‘standard condition’ control.

FIG. 21 is a plot of 3-stage cascade grinding at pilot scale in continuous mode with NBSK at 50 POP, showing the FLT-energy relationship at each stage. The square symbol is the single stage 3 mm media 50 POP ‘standard condition’ control.

FIG. 22 is a plot of 3-stage data from FIG. 21, alongside 2-stage 5 mm/1.7 mm equivalents.

FIG. 23 is a plot of FLT vs. energy input for full-scale continuous data for NBSK, 100 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes).

FIG. 24 is a plot of MFC fibre length Lc (I) vs. energy input for full-scale continuous data for NBSK, 100 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes).

FIG. 25 is a plot of FLT vs. energy input for full-scale continuous data for NBSK, 50 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes).

FIG. 26 is a plot of MFC fibre length Lc (I) vs. energy input for full-scale continuous data for NBSK, 50 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes).

FIG. 27 is a plot of FLT-energy results for pilot-scale continuous grinding with Eucalyptus (unless otherwise stated, using 100 POP and 48 MVC) comparing various routes for single stage grinding (blue), two-stage cascade grinding (red/pink), and three stage cascade grinding (grey).

FIG. 28 is a plot of MFC fibre length Lc (I) versus energy results for pilot-scale continuous grinding with Eucalyptus (unless otherwise stated, using 100 POP and 48 MVC) comparing various routes for single stage grinding (blue), two-stage cascade grinding (red/pink), and three stage cascade grinding (grey).

FIG. 29 is a plot of FLT and Lc (I) achieved by an energy input of 1500 kWh/t, comparing single stage to different cascade grind variants.

FIG. 30 is a plot of FLT and Lc (I) achieved by an energy input of 3000 kWh/t, comparing single stage to different cascade grind variants.

FIG. 31 is a plot of FLT vs. energy for full-scale 100 POP eucalyptus continuous grinds, comparing single stage to cascade grinding.

FIG. 32 is a plot of MFC fibre length Lc (I) vs. energy for full-scale 100 POP eucalyptus continuous grinds, comparing single stage to cascade grinding.

FIG. 33 is a plot of FLT-energy results for pilot-scale continuous grinding with Birch at 100 POP

FIG. 34 is a plot of MFC fibre length Lc (I) vs. energy results for pilot-scale continuous grinding with Birch at 100 POP

FIG. 35 is white light interferometer profilometry scan of (a) the smoother, 3 mm mullite formulation and (b), the rougher 3 mm mullite formulation.

FIG. 36 is a plot of media wear versus energy input for the rougher and smoother mullite media when grinding to produce MFC in the absence of mineral.

FIG. 37 is a plot of length-weighted fibre length Lc (I) versus energy input for the rougher and smoother mullite media when grinding to produce MFC in the absence of mineral.

FIG. 38 is a plot of MFC Tensile Index FLT versus energy input for the rougher and smoother mullite media when grinding to produce MFC in the absence of mineral.

FIG. 39 is a plot of media wear rate versus media surface roughness at 3000 kWh/t energy input for glass and mullite grinding media.

FIG. 40 is a plot of media wear rate versus media surface roughness at 3000 kWh/t energy input for glass, mullite, alumina/zirconium silicate, and zirconia grinding media.

FIG. 41 is a plot of MFC Lc (I) vs. media roughness for varying media compositions at 1000 kWh/t.

FIG. 42 is a plot of MFC Lc (I) vs. media roughness for varying media compositions at 3000 kWh/t.

FIG. 43 is a plot of MFC tensile index vs. mean arithmetic roughness, and shows the influence of media roughness on FLT at both 1000 kWh/t and 3000 kWh/t.

FIG. 44 is a plot of FLT tensile index versus % MOF media wear for MFC samples produced using media of various sizes, densities, and roughnesses, at various impeller speeds, at a specific energy input of 3000 kWh/t.

FIG. 45 is a plot of MFC FLT tensile index versus specific energy input for pilot batch grind energy sweeps using NBSK fibres in the absence of mineral, comparing media of various sizes and roughnesses.

FIG. 46 is a plot of media wear (% MOF) versus specific energy input for pilot batch grind energy sweeps using NBSK fibres in the absence of mineral, comparing media of various sizes and roughnesses.

FIG. 47 is a plot of Lc (I) vs. specific energy input for lab grinds of acacia.

FIG. 48 is a plot of FLT vs. specific energy input for lab grinds of acacia.

FIG. 49 is a plot showing the effect of MVC on FLT tensile index with eucalyptus fibres and smooth 1.7 mm media, and with acacia fibres and rough 3 mm media.

FIG. 50 is a plot showing the effect of MVC on media wear as % MOF with eucalyptus fibres and smooth 1.7 mm media, and with acacia fibres and rough 3 mm media.

FIG. 51 is a plot of fibre breakage factor vs. media size for acacia fibres ground to 1500 kWh/t.

FIG. 52 is a plot of fibre breakage factor vs. media size for acacia fibres ground to 3000 kWh/t.

FIG. 53 is a plot of FLT vs. media size for acacia fibres ground to 1500 kWh/t.

FIG. 54 is a plot of FLT vs. media size for acacia fibres ground to 3000 kWh/t.

FIG. 55 is a plot of FLT vs. % MOF for acacia fibres ground to 3000 kWh/t using media of different sizes (first number in label) and roughnesses (second number in label).

FIG. 56 is a plot of fibre breakage factor vs. media size for eucalyptus fibres ground to 1500 kWh/t.

FIG. 57 is a plot of fibre breakage factor vs. media size for eucalyptus fibres ground to 3000 kWh/t.

FIG. 58 is a plot of FLT vs. media size for eucalyptus fibres ground to 1500 kWh/t.

FIG. 59 is a plot of FLT vs. media size for eucalyptus fibres ground to 3000 kWh/t.

FIG. 60 is a plot of FLT vs. % MOF for eucalyptus fibres ground to 3000 kWh/t using media of different sizes (first number in label) and roughnesses (second number in label).

FIG. 61 is a plot of fibre breakage factor vs. media size for NBSK fibres ground to 1500 kWh/t.

FIG. 62 is a plot of fibre breakage factor vs. media size for NBSK fibres ground to 3000 kWh/t.

FIG. 63 is a plot of FLT vs. media size for NBSK fibres ground to 1500 kWh/t.

FIG. 64 is a plot of FLT vs. media size for NBSK fibres ground to 3000 kWh/t.

FIG. 65 is a plot of FLT vs. % MOF for NBSK fibres ground to 3000 kWh/t using media of different sizes (first number in label) and roughnesses (second number in label).

FIG. 66 is a plot of fibre breakage factor vs. media size for OCC fibres ground to 1500 kWh/t.

FIG. 67 is a plot of fibre breakage factor vs. media size for OCC fibres ground to 3000 kWh/t.

FIG. 68 is a plot of FLT vs. media size for OCC fibres ground to 1500 kWh/t.

FIG. 69 is a plot of FLT vs. media size for OCC fibres ground to 3000 kWh/t.

FIG. 70 is a plot of FLT vs. % MOF for OCC fibres ground to 3000 kWh/t using media of different sizes (first number in label) and roughnesses (second number in label).

FIG. 71 is a plot of fibre breakage factor vs. media size for kaolin-based media with roughnesses between 0.08 μm and 0.2 μm, for five fibre species.

FIG. 72 is a plot of FLT tensile index vs. media size for kaolin-based media with roughnesses between 0.08 μm and 0.2 μm, for five fibre species.

FIG. 73 is a plot of fibre breakage factor tensile index vs. media size for glass and zirconia-based media with roughnesses below 0.04 μm, for four fibre species.

FIG. 74 is a plot of FLT tensile index vs. media size for glass and zirconia-based media with roughnesses below 0.04 μm, for four fibre species.

FIG. 75 is a plot of fibre solids vs. FLT.

FIG. 76 is a plot of FLT tensile index vs. media wear content (% MOF) at an energy input of 2500 kWh/t, comparing prior art data to present work data, for four fibre species.

FIG. 77 is a plot of FLT tensile index vs. media wear content (% MOF) at an energy input of 5000 kWh/t, comparing prior art data to present work data, for four fibre species.

DETAILED DESCRIPTION OF THE INVENTION

The titles, headings and subheadings provided herein should not be interpreted as limiting the various aspects of the disclosure. Accordingly, the terms defined below are more fully defined by reference to the specification in its entirety. All references cited herein are incorporated by reference in their entirety.

The foregoing has outlined rather broadly the features and technical advantages of the present disclosure in order that the detailed description of the invention that follows may be better understood. Additional features and advantages of the invention will be described herein, which form the subject of the claims of the invention. It should be appreciated by those skilled in the art that any conception and specific embodiment disclosed herein may be readily utilized as a basis for modifying or designing other means for carrying out the same purposes of the present disclosure. It should also be realized by those skilled in the art that such equivalent means do not depart from the spirit and scope of the invention as set forth in the appended claims. The novel features which are believed to be characteristic of the invention, both as to its organization and method of operation, together with further objects and advantages, will be better understood from the following description when considered in connection with the accompanying Figures. It is to be expressly understood, however, that any description, Figure, Example, etc. is provided for the purpose of illustration and description only and is by no means intended to define the limits the invention.

Unless otherwise defined, scientific and technical terms used herein shall have the meanings that are commonly understood by those of ordinary skill in the art. Further, unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular.

In this application, the use of “or” means “and/or” unless stated otherwise. In the context of a multiple dependent claim, the use of “or” refers back to more than one preceding independent or dependent claim in the alternative only.

The use of the word “a” or “an” when used in conjunction with the term “comprising” may mean “one,” but it is also consistent with the meaning of “one or more,” “at least one,” and “one or more than one.” The use of the term “or” is used to mean “and/or” unless explicitly indicated to refer to alternatives only if the alternatives are mutually exclusive, although the disclosure supports a definition that refers to only alternatives and “and/or.”

Throughout this application, the term “about” is used to indicate that a value includes the inherent variation of error for the quantifying device, the method being employed to determine the value, or the variation that exists among the study subjects. For example, but not by way of limitation, when the term “about” is utilized, the designated value may vary by plus or minus twelve percent, or eleven percent, or ten percent, or nine percent, or eight percent, or seven percent, or six percent, or five percent, or four percent, or three percent, or two percent, or one percent. The use of the term “at least one” will be understood to include one as well as any quantity more than one, including but not limited to, 1, 2, 3, 4, 5, 10, 15, 20, 30, 40, 50, 100, etc. The term “at least one” may extend up to 100 or 1000 or more depending on the term to which it is attached. In addition, the quantities of 100/1000 are not to be considered limiting as lower or higher limits may also produce satisfactory results. In addition, the use of the term “at least one of X, Y, and Z” will be understood to include X alone, Y alone, and Z alone, as well as any combination of X, Y, and Z.

The use of ordinal number terminology (i.e., “first”, “second”, “third”, “fourth”, etc.) is solely for the purpose of differentiating between two or more items and, unless otherwise stated, is not meant to imply any sequence or order or importance to one item over another or any order of addition.

As used herein, the terms “comprising” (and any form of comprising, such as “comprise”, “comprises”, and “comprised”), “having” (and any form of having, such as “have” and “has”), “including” (and any form of including, such as “includes” and “include”), or “containing” (and any form of containing, such as “contains” and “contain”), are inclusive or open-ended and do not exclude additional, unrecited elements or method steps.

Additionally, a term that is used in conjunction with the term “comprising” is also understood to be able to be used in conjunction with the term “consisting of or “consisting essentially of.”

As used herein, the term “include” and its grammatical variants are intended to be non-limiting, such that recitation of items in a list is not to the exclusion of other like items that can be substituted or added to the listed items.

A fibrous substrate comprising cellulose (variously referred to herein as “fibrous substrate comprising cellulose,” “cellulose fibres,” “fibrous cellulose feedstock,” “cellulose feedstock” and “cellulose-containing fibres” may be derived from virgin or recycled pulp or a papermill broke and/or industrial waste, or paper streams rich in mineral fillers and cellulosic materials from a papermill.

A fibrous substrate comprising cellulose (variously referred to herein as “fibrous substrate comprising cellulose,” “cellulose fibres,” “fibrous cellulose feedstock,” “cellulose feedstock” and “cellulose-containing fibres” may be derived from non-wood cellulosic fibres. Example non-wood cellulosic fibre sources include, but are not limited to, bagasse, abaca, kenaf, sisal, cotton, hemp, flax, miscanthus, sorghum, jute, and bamboo As used herein, “FLT Index” is a tensile strength measurement performed in accordance with the procedures of Example 2.

As used herein, “mechanical properties” of MFC include one or more of the following: Tensile Strength, Tensile Elongation, Tensile Index, Burst Strength, Tear Strength, Tear Index, Scott Bond, Breaking Energy and Breaking Elongation.

As used herein, “percentage of pulp” and “POP” refer to the pulp consistency as a weight percentage of dry substances in a composition.

As used herein, the term “substantially” means that the subsequently described event or circumstance completely occurs or that the subsequently described event or circumstance occurs to a great extent or degree. For example, when associated with a particular event or circumstance, the term “substantially” means that the subsequently described event or circumstance occurs at least 80% of the time, or at least 85% of the time, or at least 90% of the time, or at least 95% of the time. Conversely, when used to signify that the mechanical properties, such as FLT tensile index and/or viscosity are “not substantially degraded” or similar language, the degradation of tensile index and/or viscosity are not diminished by more than 15%, or more than 10%, or more than 5% from the properties of the control.

As used herein, the phrase “integer from X to Y” means any integer that includes the endpoints. For example, the phrase “integer from 1 to 5” means 1, 2, 3, 4, or 5.

Microfibrillated Cellulose

Microfibrillated cellulose (MFC), although well-known and described in the art, for purposes of the presently disclosed and/or claimed inventive concept(s), microfibrillated cellulose is defined as cellulose consisting of microfibrils in the form of either isolated cellulose microfibrils and/or microfibril bundles of cellulose, both of which are derived from a cellulose raw material. Thus, microfibrillated cellulose is to be understood to comprise partly or totally fibrillated cellulose or lignocellulose fibers, which may be achieved by a variety of processes known in the art, preferably in the present application as a result of a mechanical grinding process.

As used herein, “microfibrillated cellulose” can be used interchangeably with “microfibrillar cellulose,” “nanofibrillated cellulose,” “nanofibril cellulose,” “nanofibers of cellulose,” “nanoscale fibrillated cellulose,” “microfibrils of cellulose,” and/or simply as “MFC.” Additionally, as used herein, the terms listed above that are interchangeable with “microfibrillated cellulose” may refer to cellulose that has been completely microfibrillated or cellulose that has been substantially microfibrillated but still contains an amount of non-microfibrillated cellulose at levels that do not interfere with the benefits of the microfibrillated cellulose as described and/or claimed herein.

By “microfibrillating” is meant a process in which microfibrils of cellulose are liberated or partially liberated as individual species or as small aggregates as compared to the fibres of the pre-microfibrillated pulp. Typical cellulose fibres (i.e., pre-microfibrillated pulp) suitable for use in papermaking include larger aggregates of hundreds or thousands of individual cellulose fibrils.

Microfibrillated cellulose comprises cellulose, which is a naturally occurring polymer comprising repeated glucose units. The term “microfibrillated cellulose”, also denoted MFC, as used in this specification, includes microfibrillated/microfibrillar cellulose and nano-fibrillated/nanofibrillar cellulose (NFC), which materials are also sometimes called nanocellulose.

Microfibrillated cellulose may also be prepared by stripping away the outer layers of cellulose fibers that may have been exposed through mechanical shearing, with or without prior enzymatic or chemical treatment.

In a non-limiting example, the term microfibrillated cellulose is used to describe fibrillated cellulose comprising nanoscale cellulose particle fibers or fibrils frequently having at least one dimension less than 100 nm. When liberated from cellulose fibres, fibrils typically have a diameter less than 100 nm. The actual diameter of cellulose fibrils depends on the source and the method of measuring such fibrils as well as the manufacturing methods that are employed.

The particle size distribution and/or aspect ratio (length/width) of the cellulose microfibrils attached to the fibrillated cellulose fibre or as a liberated microfibril depends on the source and the manufacturing methods employed in the microfibrillation process.

In a non-limiting example, the aspect ratio of microfibrils is typically high and the length of individual microfibrils may be more than one micrometer and the diameter may be within a range of about 5 to 60 nm with a number-average diameter typically less than 20 nm. The diameter of microfibril bundles may be larger than 1 micron.

In a non-limiting example, the smallest fibril is conventionally referred to as an elementary fibril, which generally has a diameter of approximately 2-4 nm. It is also common for elementary fibrils to aggregate, which may also be considered as microfibrils.

In a non-limiting example, the microfibrillated cellulose may at least partially comprise nanocellulose. The nanocellulose may comprise mainly nano-sized fibrils having a diameter that is less than 100 nm and a length that may be in the micron-range or lower. The smallest microfibrils are like so-called elementary fibrils, the diameter of which is typically 2 to 4 nm. Of course, the dimensions and structures of microfibrils and microfibril bundles depend on the raw materials used in addition to the methods of producing the microfibrillated cellulose. Nonetheless, it is expected that a person of ordinary skill in the art would understand the meaning of “microfibrillated cellulose” in the context of the presently disclosed and/or claimed inventive concept(s).

Depending on the source of the cellulose fibres and the manufacturing process employed to microfibrillate the cellulose fibres, the length of the fibrils can vary, frequently from about 1 to greater than 10 micrometres.

A coarse MFC grade might contain a substantial fraction of fibrillated fibres, i.e. protruding fibrils from the tracheid (cellulose fibre), and with a certain amount of fibrils liberated from the tracheid (cellulose fibre).

In an embodiment, the microfibrillated cellulose may also be prepared from recycled pulp or a papermill broke and/or industrial waste, or a paper stream rich in mineral fillers and cellulosic materials from a papermill.

In an embodiment, the microfibrillated cellulose may also be prepared from non-wood cellulosic fibres (e.g., those originating from bagasse, abaca, kenaf, sisal, cotton, hemp, flax, miscanthus, sorghum, jute, bamboo, and mixtures thereof).

The fibrous substrate comprising cellulose may be added to a grinding vessel in a dry state. For example, a dry paper broke may be added directly to the grinder vessel. The aqueous environment in the grinder vessel will then facilitate the formation of a pulp.

Generally, the present invention is related to modifications, for example, improvements, to the methods and compositions described in WO2010/131016, WO2015/173376 and WO2018193314, the entire contents of which are hereby incorporated by reference.

Co-Processing of a Fibrous Substrate Comprising Cellulose and at Least One Inorganic Particulate Material

WO2010/131016 discloses a process for preparing microfibrillated cellulose comprising microfibrillating, e.g., by grinding a fibrous material comprising cellulose, optionally in the presence of grinding medium and/or inorganic particulate material. When used as a filler in paper, for example, as a replacement or partial replacement for a conventional mineral filler, the microfibrillated cellulose obtained by said process, optionally in combination with inorganic particulate material, was found to improve the burst strength properties of the paper. That is, relative to a paper filled with exclusively mineral filler, paper filled with the microfibrillated cellulose was found to have improved burst strength. In other words, the microfibrillated cellulose filler was found to have paper burst strength enhancing attributes. In one particularly advantageous embodiment of that invention, the fibrous material comprising cellulose was ground in the presence of a grinding medium, optionally in combination with inorganic particulate material, to obtain microfibrillated cellulose having a fibre steepness of from about 20 to about 50.

As used herein, the terms “co-grinding (or “co-ground”) composite or composition comprising microfibrillated cellulose and inorganic particulate material” refers to a composite or composition obtained by a “co-grinding microfibrillation process,” wherein a fibrous substrate comprising cellulose is microfibrillated in an aqueous environment in a grinding apparatus in the presence of the at least one inorganic particulate material, and optionally a grinding medium other than the at least one inorganic particulate material (or stated differently by “co-processing” a fibrous substrate comprising cellulose in the presence of the at least one inorganic particulate material in a wet stirred media mill and optionally in the presence of a grinding medium other than the at least one inorganic particulate material, which is removed after grinding, to produce microfibrillated cellulose). See the description below of exemplary microfibrillation processes and wet-grinding processes.

After co-processing to form a co-processed microfibrillated cellulose and inorganic particulate material composite, additional inorganic particulate material may be added (e.g., by blending or mixing) to reduce the microfibrillated cellulose content of the co-processed microfibrillated cellulose and inorganic particulate material composite.

In an embodiment, the MFC may be manufactured using a wet stirred media mill, such as a tower mill, a horizontal milling apparatus (e.g., an IsaMill as manufactured by Glencore Technology, Sala agitated mill (SAM) (for example, manufactured by Svedala AB), a screened grinding mill, or a stirred media detritor.

A wet stirred media mill consists of a rotating impeller that transfers kinetic energy to small grinding media beads, which grind down the charge via a combination of shear, compressive, and impact forces. A variety of grinding apparatus may be used to produce MFC by the disclosed methods herein, including, for example, a tower mill, a screened grinding mill, an ISA mil, Sala agitated mill (SAM), or a stirred media detritor (SMD).

Hardwood and Softwood Pulps

“Hardwood pulp fibre” as used herein means pulp fibres obtained from deciduous trees. Non-limiting examples of deciduous trees include Northern hardwood trees and tropical hardwood trees. Non-limiting examples of hardwood pulp fibres include hardwood pulp fibres obtained from a fibre source selected from the group consisting of Acacia, Eucalyptus, Maple, Oak, Aspen, Birch, Cottonwood, Alder, Ash, Cherry, Elm, Hickory, Poplar, Gum, Walnut, Locust, Sycamore, Beech, Catalpa, Sassafras, Gmelina, Albizia, Androcephalous, Magnolia, and mixtures thereof. In one example, the hardwood pulp fibre of the present invention is obtained from Eucalyptus.

Wood pulps utilized in the paper and board industry come in a variety of forms from commercial pulp manufacturing companies. These pulp forms include mechanical pulp, thermomechanical pulp, chemi-thermomechanical pulp, and chemical pulp, including, for example, Northern Bleached Softwood Kraft pulp (“NSBK”), Bleached Softwood Kraft pulp, Bleached Hardwood pulp, unbleached softwood and hardwood Kraft pulps, Sulphite Bleached pulp, Bleached Chemi-Thermo Mechanical Pulp (“BCTMP”), and recycled pulp.

Tropical hardwood pulp fibre as used herein means pulp fibres obtained from a tropical hardwood tree. Non-limiting examples of tropical hardwood trees include Eucalyptus trees and/or Acacia trees.

Pulp may be obtained from many wood sources, which typically are classified into two classes, namely softwood and hardwood pulps. Softwood pulps are favored in many applications, since the cellulose fibres are typically longer. Softwood pulps may be processed from spruce, pine, fir, larch, and hemlock, whereas hardwood pulps are typically processed from eucalyptus, aspen and birch, for example.

Pulps are processed conventionally from wood chips into pulp sheets that are shipped, for example, to paper mills for processing into paper and paperboards. Sawmill residue chips from sapwood are typically used for Kraft chemical processing, but also whole-log wood chips may be used, which in addition to sapwood contain heartwood from the wood logs. Sapwood is favoured for chemical processing, since sapwood has fibres with less lignin, lower density, less wood extractives, less acidic, higher moisture content and more living cells; and, consequently is easier to cook. Heartwood is more difficult to penetrate with cooking liquors than sapwood.

Regardless of the source of wood chips, and especially if multiple sources of wood chips are pulped together, the resulting wood pulp produced by the Kraft process has numerous variables, including, for example, rejects, bark content, moisture content, Kappa number variability, biological knots, decayed wood, sulfidity percentage and numerous other variables. These variables are impacted by the type of pulping process utilized. The objective of Kraft processing is to provide uniform delignification and high cooking yield and pulp quality.

Wood chips processed into wood pulps have three main components, apart from water, which are cellulose fibres, lignins and hemicelluloses. These three components are profoundly differentiated based on the type of processing employed to pulp the wood chips from softwood or hardwood sources.

One of the most commercially significant wood pulps available for a variety of end-use applications is Northern Bleached Softwood Kraft pulp, or NBSK. The commercially available NBSK pulp comprises long slender cellulose-containing fibres that provide excellent bonding and tensile properties. NBSK pulp is conventionally used for manufacturing a variety of paper products, including printing, and writing paper, specialty grades, and a range of tissue products.

Wet Grinding Process Without Inorganic Particulate Material

The present disclosure provides, in some embodiments, microfibrillating a fibrous substrate comprising cellulose by grinding in the presence of a particulate grinding medium which is to be removed after the completion of grinding. By microfibrillating the cellulose, particular characteristics, and properties, including the characteristics and properties described herein, are imparted to the microfibrillated cellulose and the compositions comprising the microfibrillated cellulose.

Various methods of producing microfibrillated cellulose (“MFC”) are known in the art. Certain methods and compositions comprising microfibrillated cellulose produced by grinding procedures are described in WO-A-2010/131016. Husband, J. C., Svending, P., Skuse, D. R., Motsi, T., Likitalo, M., Coles, A., FiberLean Technologies Ltd., 2015, “Paper filler composition,” PCT International Application No. WO2010/131016, the contents of which is hereby incorporated by reference in its entirety. Paper products comprising such microfibrillated cellulose have been shown to exhibit excellent paper properties, such as paper burst and tensile strength. The methods described in WO2010/131016 also enable the production of microfibrillated cellulose economically.

WO2010/131016 describes a grinding procedure to produce microfibrillated cellulose with or without inorganic particulate material. Such a grinding procedure is described below. In an embodiment of the process set forth in WO2010/131016, the process utilizes mechanical disintegration of cellulose fibres to produce microfibrillated cellulose (“MFC”) cost-effectively and at large scale, without requiring cellulose pre-treatment. An embodiment of the method uses stirred media detritor grinding technology, which disintegrates fibres into MFC by agitating grinding media particles in the form of beads. In this process, a mineral such as calcium carbonate or kaolin may optionally be added as a grinding aid, greatly reducing the energy required. Husband, J. C., Svending, P., Skuse, D. R., Motsi, T., Likitalo, M., Coles, A., FiberLean Technologies Ltd., 2015, “Paper filler composition,” U.S. Pat. U.S. Pat. No. 9,127,405B2, the contents of which is hereby incorporated by reference in its entirety.

Grinding Media

The grinding media may comprise mullite (i.e., a media produced by calcining clay-based particles), alumina, silicate, zirconia, zirconium silicate, or a combination thereof and has a specific gravity of at least about 2.5, for example, at least about 3, or at least about 3.5, or at least about 4.0, or at least about 4.5, or least about 5.0, or at least about 5.5, or at least about 6.0. The grinding media comprises particles having an average diameter in the range of from about 1 mm to about 3 mm. the grinding media size may be varied between successive vessels in the cascade to reduce grinding of the inorganic particulate material and to target grinding of the fibrous substrate comprising cellulose. The grinding medium has an arithmetic surface roughness from about 0.1 μm to about 1 μm.

By “particulate grinding medium” is meant a medium other than the inorganic particulate material which, in certain embodiments, is co-ground with the fibrous substrate comprising cellulose. Advantageously, it has been found that a particulate grinding medium having a relatively rough surface facilitates, e.g., enhances, the production of microfibrils during the manufacture of microfibrillated cellulose. It is believed that microfibrils are formed due to the intimate interaction of the particulate grinding media surface which has a relatively rough texture and cellulose fibres during the grinding process. Without wishing to be bound by theory, it is thought that the mechanism of microfibril production is due to the relatively rough surface of the particulate grinding media results in ‘hooking’ and ‘tearing’ and/or ‘delayering’ cellulose during grinding. The interaction between the particulate grinding media and cellulose that results in microfibrillated cellulose may include media-cellulose collisions, shear of cellulose between media particulates or between media particulates and grinder wall.

Surface roughness may be determined by optical interferometry, i.e., the measurement of the surface topography of a test surface of the particulate grinding medium relative to a reference surface, as carried out by an optical interferometer. In certain embodiments, surface roughness is determined in accordance with the following method. A representative sample of the particulate grinding medium is obtained and placed in an interferometer coupled to an optical microscope. A suitable interferometer is an Omniscan MicroXAM2. A suitable optical microscope is a Keyence Optical Microscope. A representative sample consists of 5 individual particles (e.g., beads) of the particulate grinding medium to be analysed, selected at random from any given batch of particulate grinding medium. A surface roughness for each individual particle is determined at two different locations on the surface, and the 10 results (i.e., two per particle) averaged. The size of the surface area analysed at each location on each particle is constant. A suitable interferometer operating procedure is provided in Example 1. In certain embodiments, surface roughness is determined in accordance with the interferometer operating procedure provided in Appendix 1, or any other suitable procedure which provides essentially the same result.

Mean coefficient of friction may be determined by tribometry, i.e., the measurement of friction on a surface, as carried out with a tribometer. A tribometer measures the magnitude of friction and wear as surfaces are rubbed over each other. In certain embodiments, mean coefficient of friction is determined in accordance with the following method. Three individual specimens (e.g., beads) of the particulate grinding medium to be analysed are obtained, and each specimen subjected to three identical runs in a tribometer. A friction coefficient is determined for each run, giving nine friction coefficient measurements (i.e., three for each specimen). A mean coefficient of friction is obtained by adding together the nine friction coefficient measurements and dividing by nine. A suitable tribometer operating procedure is provided in Example 2. In certain embodiments, mean coefficient of friction is determined in accordance with the tribometer operating procedure provided in Example 2, or any other suitable procedure which produces essentially the same result.

The particulate grinding media may comprise particles of any suitable shape, e.g., balls, beads, cylpebs, pellets, rods, discs, cubes, toroids, cones, and the like.

In certain embodiments, the particulate grinding media comprises substantially spherical particles, e.g., balls and/or beads. For example, the grinding media may comprise at least 10% by weight of substantially spherical particles, or may comprise at least 20% by weight of substantially spherical particles, or may comprise at least 30% by weight of substantially spherical particles, or may comprise at least 40% by weight of substantially spherical particles, or may comprise at least 50% by weight of substantially spherical particles, or may comprise at least 60% by weight of substantially spherical particles, or may comprise at least 70% by weight substantially spherical particles, or may comprise at least 80% by weight of substantially spherical particles, or may comprise at least 90% by weight of substantially spherical particles, or may comprise essentially only (e.g., 95% by weight or more, or at least 99% by weight) substantially spherical particles.

In certain embodiments, the grinding medium comprises rod-shaped particles, for example, rod-shaped particles having an aspect ratio of equal to or greater than about 2:1.

The rod-shaped particles are solid bodies which have an axis running the length of the body about which an outer surface is defined, and opposite end surfaces. The outer surface and the opposite end surfaces together define the body. In certain embodiments, the lengthwise axis is substantially rectilinear, by which we mean that the line representing the shortest distance between the two ends falls completely within the body. In other embodiments, the rod-shaped particles may take an arcuate form in which the axis is curvilinear and the line representing the shortest distance does not fall completely within the body. Mixtures of rod-shaped bodies having a rectilinear axis and bodies having an arcuate form are contemplated, as are embodiments in which substantially all (for example 90% by weight or 95% by weight or 99% by weight) of the rod-shaped particles of aspect ratio of 2:1 or more either have the rectilinear form or have the arcuate form.

In certain embodiments, the cross section of the rod-shaped particles is substantially constant along the length of the particle. By “substantially constant” is meant that the major dimension of the cross-section does not vary by, for example, more than 20% or by more than 10% or by more than 5%. In another embodiment, the cross-section of the rod-shaped particles varies along the length of the particle by, for example, by more than 20%. For example, the body of the rod-shaped particle may take the form of a barrel in which the cross-section at each of the ends of the body of the particle is less than a cross-section measured between the ends; or for example, the body of the rod-shaped particle may take the form of an inverse barrel in which the cross-section at each of the ends of the particle is greater than a cross-section measured between the ends. The cross-sectional shape of the rod-shaped particles may be symmetrical or asymmetrical. For example, the cross-sectional shape may be circular or substantially circular, or may be substantially ovoid. Other shapes include angular shapes such as triangles, squares, rectangles, stars (five and six-pointed), diamonds, etc. The boundary between the outer lengthwise surface and the opposite end surfaces may be angular, i.e. having a discrete sharp boundary, or non-angular, i.e. being rounded or radiused. The end surfaces may be flat, convex or concave.

As previously noted, the aspect ratio of the rod-shaped particles is advantageously 2:1 or more than 2:1. The aspect ratio is to be understood as the ratio of the longest dimension of the particle to the shortest dimension. For present purposes, the longest dimension is the axial length of the rod-shaped particles. Where the particle has a constant cross-section along its length, the shortest dimension for purposes of defining the aspect ratio is the largest dimension of the cross-section which passes through the geometric centre of the particle cross-section. Where the cross-section varies along the length of the particle, the shortest dimension for purposes of defining the aspect ratio is the largest dimension at the point at which the cross-section is at a maximum. Where the particle has an irregular shaped cross-section, the shortest dimension for the purposes of defining the aspect ratio is the maximum transverse dimension perpendicular to the axis of the rod-shaped particle. An example of suitable rod-shaped particles for use in certain embodiments of the invention are particles having a substantially rectilinear axis and a substantially circular cross section.

Another example of suitable rod-shaped particles for use in certain embodiments of the invention are particles having a arcuate form and a substantially circular cross-section. In both these examples, the boundary between the outer lengthwise surface and the opposite end surfaces is rounded and the ends are generally flat or convex. In certain embodiments, the rod-shaped particles have an aspect ratio of 2.5:1 or more than 2.5:1, or an aspect ratio of 3:1 or more than 3:1, or an aspect ratio of 4:1 or more than 4:1, or an aspect ratio of 5:1 or more than 5:1, or an aspect ratio of 6:1 or more than 6:1. The aspect ratio may be 10:1 or less than 10:1, or may be 9:1 or less than 9:1 or may be 8:1 or less than 8:1 or may be 7:1 or less than 7: or may be 6:1 or less than 6:1 or may be 5:1 or less than 5:1. The aspect ratio may be in the range of from 2:1 to 10:1 or may be in the range of from 2:1 to 5:1 or may be in the range 3:1 to 8:1 or may be in the range of from 3:1 to 6:1.

In certain embodiments, the axial length of the rod-shaped particles ranges from about 1 mm to about 5 mm, or from about 2 mm to about 4 mm, or from about 1 mm to about 3 mm. In another embodiment, the rod length is less than about 3 mm.

In certain embodiments, the grinding media may comprise (i.e., in addition to the rod-shaped particles having an aspect ratio of 2:1 or more) other particles selected from rod-shaped particles having an aspect ratio less than 2:1 and particles having other shapes such as spheres, cylpebs, cubes, discs, toroids, cones, and the like. For example, the grinding media may comprise at least 10% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 20% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 30% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 40% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 50% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 60% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 70% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 80% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise at least 90% by weight of rod-shaped particles having an aspect ratio of 2:1 or more, or may comprise essentially only (e.g. 95% by weight or more) rod-shaped particles having an aspect ratio of 2:1 or more. It will be further understood that in certain embodiments of the invention, a relatively small number of shaped particles having an aspect ratio smaller than 2:1 may be present as a by-product of the process by which the particles are made or handled. Similarly, rod-shaped particles having a relatively high aspect ratio such as, for example, greater than about 10:1, may be added to the grinding process, in which case these rods may snap to their own preferred length during the grinding process. It will also be understood that as the grinding process progresses the shape of at least some of the rod-shaped particles may evolve such that the ends round off, and the aspect ratio lowers, and in some cases the virgin rod-shaped particles may eventually become small spheres, so a typical mature grinder may contain rods, worn rods and even spheres. Thus, a “worked-in” sample of rod-shaped particles which originally had an aspect ratio at least 2:1 or more may contain a majority (if worked long enough) of particles somewhat different in shape to the rod-shaped particles comprised in the virgin media. The grinder may be topped up with fresh media comprising rod-shaped particles having an aspect ratio of 2:1 or more.

The fibrous substrate comprising cellulose may be derived from any suitable source, such as wood, grasses (e.g., sugarcane, bamboo) or rags (e.g., textile waste, cotton, hemp or flax). The fibrous substrate comprising cellulose may be in the form of a pulp (i.e., a suspension of cellulose fibres in water), which may be prepared by any suitable chemical or mechanical treatment, or combination thereof. For example, the pulp may be a chemical pulp, or a chemithermomechanical pulp, or a mechanical pulp, or a recycled pulp, or a papermill broke, or a papermill waste stream, or waste from a papermill, or a combination thereof. The cellulose pulp may be beaten (for example in a Valley beater) and/or otherwise refined (for example, processing in a conical or plate refiner) to any predetermined freeness, reported in the art as Canadian standard freeness (CSF) in cm3. CSF means a value for the freeness or drainage rate of pulp measured by the rate that a suspension of pulp may be drained. For example, the cellulose pulp may have a Canadian standard freeness of about 10 cm3 or greater prior to being microfibrillated. The cellulose pulp may have a CSF of about 700 cm3 or less, for example, equal to or less than about 650 cm3, or equal to or less than about 600 cm3, or equal to or less than about 550 cm3, or equal to or less than about 500 cm3, or equal to or less than about 450 cm3, or equal to or less than about 400 cm3, or equal to or less than about 350 cm3, or equal to or less than about 300 cm3, or equal to or less than about 250 cm3, or equal to or less than about 200 cm3, or equal to or less than about 150 cm3, or equal to or less than about 100 cm3, or equal to or less than about 50 cm3. The cellulose pulp may then be dewatered by methods well known in the art, for example, the pulp may be filtered through a screen in order to obtain a wet sheet comprising at least about 10% solids, for example at least about 15% solids, or at least about 20% solids, or at least about 30% solids, or at least about 40% solids. The pulp may be utilised in an unrefined state, that is to say without being beaten or dewatered, or otherwise refined.

The step of microfibrillating may be carried out in any suitable apparatus, including but not limited to a refiner. In one embodiment, the microfibrillating step is conducted in a grinding vessel. The microfibrillated step may be carried out in an aqueous environment, i.e., under wet-grinding conditions. In another embodiment, the microfibrillating step is carried out in a homogenizer.

In certain embodiments, the microfibrillating process, e.g., grinding, is carried out in the presence of grindable inorganic particulate material. In certain embodiments, the grinding is carried out in the absence of grindable inorganic particulate material.

The grinding medium may be present in an amount up to about 70% by volume of the charge. The grinding media may be present in amount of at least about 10% by volume of the charge, for example, at least about 20% by volume of the charge, or at least about 30% by volume of the charge, or at least about 40% by volume of the charge, or at least about 50% by volume of the charge, or at least about 60% by volume of the charge. In certain embodiments, the grinding medium is present in an amount from about 30 to about 70% by volume of the charged, for example, from about 40 to about 60% by volume of the charge, for example, from about 45 to about 55% by volume of the charge.

By ‘charge’ is meant the composition which is the feed fed to the grinder vessel. The charge includes water (when present), grinding media, fibrous substrate comprising cellulose and inorganic particulate material (when present), and any other optional additives (when present) as described herein.

The grinding may be performed in a vertical mill or a horizontal mill.

In certain embodiments, the grinding is performed in a grinding vessel, such as a tumbling mill (e.g., rod, ball and autogenous), a stirred mill (e.g., SAM or IsaMill), a tower mill, a stirred media detritor (SMD), or a grinding vessel comprising rotating parallel grinding plates between which the feed to be ground is fed.

In one embodiment, the grinding vessel is a vertical mill, for example, a stirred mill, or a stirred media detritor, or a tower mill.

The vertical mill may comprise a screen above one or more grind zones. In an embodiment, a screen is located adjacent to a quiescent zone and/or a classifier. The screen may be sized to separate grinding media from the product aqueous suspension comprising microfibrillated cellulose and inorganic particulate material and to enhance grinding media sedimentation.

In one embodiment, the grinding vessel is a tower mill. The tower mill may comprise a quiescent zone above one or more grinding zones. A quiescent zone is a region located towards the top of the interior of tower mill in which minimal or no grinding takes place and comprises microfibrillated cellulose and (when present) inorganic particulate material. The quiescent zone is a region in which particles of the grinding medium sediment down into the one or more grinding zones of the tower mill.

The tower mill may comprise a vertical impeller shaft equipped with a series of impeller rotor disks throughout its length. The action of the impeller rotor disks creates a series of discrete grinding zones throughout the mill.

The tower mill may comprise a classifier above one or more grinding zones. In an embodiment, the classifier is top mounted and located adjacent to a quiescent zone. The classifier may be a hydrocyclone.

The tower mill may comprise a screen above one or more grind zones. In an embodiment, a screen is located adjacent to a quiescent zone and/or a classifier. The screen may be sized to separate grinding media from the product aqueous suspension comprising microfibrillated cellulose and (when present) inorganic particulate material and to enhance grinding media sedimentation.

In another embodiment, the grinding is performed in a screened grinder, for example, a stirred media detritor. The screened grinder may comprise one or more screen(s) sized to separate grinding media from the product aqueous suspension comprising microfibrillated cellulose and inorganic particulate material. The screened grinder may comprise one or more screen(s) having a nominal aperture size of at least about 250 μm, for example, the one or more screens may have a nominal aperture size of at least about 300 μm, or at least about 350 μm, or at least about 400 μm, or at least about 450 μm, or at least about 500 μm, or at least about 550 μm, or at least about 600 μm, or at least about 650 μm, or at least about 700 μm, or at least about 750 μm, or at least about 800 μm, or at least about 850 μm, or at or least about 900 μm, or at least about 1000 μm, or at least about 1250 μm, or at least about 1500 μm. In certain embodiments, the screened grinder may comprise one or more screen(s) having a nominal aperture size of up to about 4000 μm, for example, up to about 3500 μm, or up to about 3000 μm, or up to about 2500 μm, or up to about 2000 μm.

In certain embodiments, at least about 5% by weight of the initial solids content may be fibrous substrate comprising cellulose, for example, at least about 10%, or at least about 15%, or at least about 20% by weight of the initial solids content may be fibrous substrate comprising cellulose.

As such, the present inventors have surprisingly found that a cellulose pulp can be microfibrillated at relatively lower energy input when it is ground in the presence of particulate grinding medium having (i) an arithmetic surface roughness of about 0.02 μm to about 2 μm, or (ii) a mean coefficient of friction of at least about 0.10, or both (i) and (ii). In other words, the particulate grinding medium may be used in order to reducing the energy input per unit amount of microfibrillated cellulose produced. Further, as described above, in certain embodiments, the use of a particulate grinding medium having i) a surface roughness of at least about 0.5 μm, or (ii) a mean coefficient of friction of at least about 0.10, or both (i) and (ii) may improve one or more properties of the microfibrillated cellulose, e.g., a strength property of the microfibrillated cellulose and/or paper products (e.g., burst strength) comprising the microfibrillated cellulose.

When present, the inorganic particulate material may, for example, be an alkaline earth metal carbonate or sulphate, such as calcium carbonate, magnesium carbonate, dolomite, gypsum, a hydrous kandite clay such as kaolin, halloysite or ball clay, an anhydrous (calcined) kandite clay such as metakaolin or fully calcined kaolin, talc, mica, perlite or diatomaceous earth, or magnesium hydroxide, or aluminium trihydrate, or combinations thereof.

In certain embodiments, the inorganic particulate material comprises or is calcium carbonate. Hereafter, certain embodiments of the invention may tend to be discussed in terms of calcium carbonate, and in relation to aspects where the calcium carbonate is processed and/or treated. The invention should not be construed as being limited to such embodiments.

The particulate calcium carbonate used in certain embodiments of the present invention may be obtained from a natural source by grinding. Ground calcium carbonate (GCC) is typically obtained by crushing and then grinding a mineral source such as chalk, marble or limestone, which may be followed by a particle size classification step, in order to obtain a product having the desired degree of fineness. Other techniques such as bleaching, flotation and magnetic separation may also be used to obtain a product having the desired degree of fineness and/or colour. The particulate solid material may be ground autogenously, i.e. by attrition between the particles of the solid material themselves, or, alternatively, in the presence of a particulate grinding medium comprising particles of a different material from the calcium carbonate to be ground. These processes may be carried out with or without the presence of a dispersant and biocides, which may be added at any stage of the process.

Precipitated calcium carbonate (PCC) may be used as the source of particulate calcium carbonate in certain embodiments of the present invention, and may be produced by any of the known methods available in the art. TAPPI Monograph Series No 30, “Paper Coating Pigment”, pages 34-35 describes the three main commercial processes for preparing precipitated calcium carbonate which is suitable for use in preparing products for use in the paper industry, but may also be used in the practice of certain embodiments of the present invention. In all three processes, a calcium carbonate feed material, such as limestone, is first calcined to produce quicklime, and the quicklime is then slaked in water to yield calcium hydroxide or milk of lime. In the first process, the milk of lime is directly carbonated with carbon dioxide gas. This process has the advantage that no by-product is formed, and it is relatively easy to control the properties and purity of the calcium carbonate product. In the second process the milk of lime is contacted with soda ash to produce, by double decomposition, a precipitate of calcium carbonate and a solution of sodium hydroxide. The sodium hydroxide may be substantially completely separated from the calcium carbonate if this process is used commercially. In the third main commercial process the milk of lime is first contacted with ammonium chloride to give a calcium chloride solution and ammonia gas. The calcium chloride solution is then contacted with soda ash to produce by double decomposition precipitated calcium carbonate and a solution of sodium chloride. The crystals can be produced in a variety of different shapes and sizes, depending on the specific reaction process that is used. The three main forms of PCC crystals are aragonite, rhombohedral and scalenohedral, all of which are suitable for use in certain embodiments of the present invention, including mixtures thereof.

Wet grinding of calcium carbonate involves the formation of an aqueous suspension of the calcium carbonate which may then be ground, optionally in the presence of a suitable dispersing agent. Reference may be made to, for example, EP-A-614948 (the contents of which are incorporated by reference in their entirety) for more information regarding the wet grinding of calcium carbonate.

In some circumstances, minor additions of other minerals may be included, for example, one or more of kaolin, calcined kaolin, wollastonite, bauxite, talc or mica, could also be present.

When the inorganic particulate material is obtained from naturally occurring sources, it may be that some mineral impurities will contaminate the ground material. For example, naturally occurring calcium carbonate can be present in association with other minerals. Thus, in some embodiments, the inorganic particulate material includes an amount of impurities. In general, however, the inorganic particulate material used in certain embodiments of the invention will contain less than about 5% by weight, preferably less than about 1% by weight, of other mineral impurities.

The inorganic particulate material may have a particle size distribution such that at least about 10% by weight, for example at least about 20% by weight, for example at least about 30% by weight, for example at least about 40% by weight, for example at least about 50% by weight, for example at least about 60% by weight, for example at least about 70% by weight, for example at least about 80% by weight, for example at least about 90% by weight, for example at least about 95% by weight, or for example about 100% of the particles have an e.s.d of less than 2 μm.

In certain embodiments, at least about 50% by weight of the particles have an e.s.d of less than 2 μm, for example, at least about 55% by weight of the particles have an e.s.d of less than 2 μm, or at least about 60% by weight of the particles have an e.s.d of less than 2 μm

Unless otherwise stated, particle size properties referred to herein for the inorganic particulate materials are as measured in a well known manner by sedimentation of the particulate material in a fully dispersed condition in an aqueous medium using a Sedigraph 5100 machine as supplied by Micromeritics Instruments Corporation, Norcross, Georgia, USA (web-site: www.micromeritics.com), referred to herein as a “Micromeritics Sedigraph 5100 unit”. Such a machine provides measurements and a plot of the cumulative percentage by weight of particles having a size, referred to in the art as the ‘equivalent spherical diameter’ (e.s.d), less than given e.s.d values. The mean particle size d50 is the value determined in this way of the particle e.s.d at which there are 50% by weight of the particles which have an equivalent spherical diameter less than that d50 value.

Alternatively, where stated, the particle size properties referred to herein for the inorganic particulate materials are as measured by the well known conventional method employed in the art of laser light scattering, using a Malvern Mastersizer S machine as supplied by Malvern Instruments Ltd (or by other methods which give essentially the same result). In the laser light scattering technique, the size of particles in powders, suspensions and emulsions may be measured using the diffraction of a laser beam, based on an application of Mie theory. Such a machine provides measurements and a plot of the cumulative percentage by volume of particles having a size, referred to in the art as the ‘equivalent spherical diameter’ (e.s.d), less than given e.s.d values. The mean particle size d50 is the value determined in this way of the particle e.s.d at which there are 50% by volume of the particles which have an equivalent spherical diameter less than that d50 value.

Thus, in another embodiment, the inorganic particulate material may have a particle size distribution, as measured by the well known conventional method employed in the art of laser light scattering, such that at least about 10% by volume, for example at least about 20% by volume, for example at least about 30% by volume, for example at least about 40% by volume, for example at least about 50% by volume, for example at least about 60% by volume, for example at least about 70% by volume, for example at least about 80% by volume, for example at least about 90% by volume, for example at least about 95% by volume, or for example about 100% by volume of the particles have an e.s.d of less than 2 μm.

In certain embodiments, at least about 50% by volume of the particles have an e.s.d of less than 2 μm, for example, at least about 55% by volume of the particles have an e.s.d of less than 2 μm, or at least about 60% by volume of the particles have an e.s.d of less than 2 μm

Details of the procedure that may be used to characterise the particle size distributions of mixtures of inorganic particle material and microfibrillated cellulose using the well known conventional method employed in the art of laser light scattering are provided in WO-A-2010/131016 at page 40, line 32 to page 41, line 34, the entire contents of which are hereby incorporated by reference.

In some embodiments, the inorganic particulate material is kaolin clay. Hereafter, this section of the specification may tend to be discussed in terms of kaolin, and in relation to aspects where the kaolin is processed and/or treated. The invention should not be construed as being limited to such embodiments. Thus, in some embodiments, kaolin is used in an unprocessed form.

Kaolin clay used in certain embodiments of this invention may be a processed material derived from a natural source, namely raw natural kaolin clay mineral. The processed kaolin clay may typically contain at least about 50% by weight kaolinite. For example, most commercially processed kaolin clays contain greater than about 75% by weight kaolinite and may contain greater than about 90%, in some cases greater than about 95% by weight of kaolinite.

Kaolin clay used in the present invention may be prepared from the raw natural kaolin clay mineral by one or more other processes which are well known to those skilled in the art, for example by known refining or beneficiation steps.

For example, the clay mineral may be bleached with a reductive bleaching agent, such as sodium hydrosulfite. If sodium hydrosulfite is used, the bleached clay mineral may optionally be dewatered, and optionally washed and again optionally dewatered, after the sodium hydrosulfite bleaching step.

The clay mineral may be treated to remove impurities, e.g. by flocculation, flotation, or magnetic separation techniques well known in the art. Alternatively the clay mineral may be untreated in the form of a solid or as an aqueous suspension.

The process for preparing the particulate kaolin clay used in certain embodiments of the present invention may also include one or more comminution steps, e.g., grinding or milling. Light comminution of a coarse kaolin is used to give suitable delamination thereof. The comminution may be carried out by use of beads or granules of a plastic (e.g. nylon), sand or ceramic grinding or milling aid. The coarse kaolin may be refined to remove impurities and improve physical properties using well known procedures. The kaolin clay may be treated by a known particle size classification procedure, e.g., screening and centrifuging (or both), to obtain particles having a desired d50 value or particle size distribution.

The relative amounts of inorganic particulate material and cellulosic material, including microfibrillated cellulose, may vary in a ratio of from about 99.5:0.5 to about 0.5:99.5, based on the dry weight of inorganic particulate material and cellulosic material, for example, a ratio of from about 99.5:0.5 to about 50:50 based on the dry weight of inorganic particulate material and cellulosic material. For example, the ratio of the amount of inorganic particulate material and cellulosic material may be from about 99.5:0.5 to about 70:30. In certain embodiments, the ratio of inorganic particulate material to cellulosic material is about 80:20, or for example, about 85:15, or about 90:10, or about 91:9, or about 92:8, or about 93:7, or about 94:6, or about 95:5, or about 96:4, or about 97:3, or about 98:2, or about 99:1. In certain embodiment, the microfibrillated cellulose obtainable by the present disclosure comprises up to about 80% by weight water, for example, up to about 75% water, or up to about 70%, or up to about 65% by weight water, or up to about 60% by weight water, or up to about 55% by weight water, or up to about 50% by weight water, or up to about 45% by weight water, or up to about 40% by weight water, or up to about 35% by weight water, or up to about 30% by weight water, or up to about 25% by weight water.

In certain embodiments, microfibrillated cellulose obtainable by the present disclosure comprises from about 50 to about 70% by weight water, for example, from about 55 to about 65% by weight water, or from about 60 to about 70% by weight water, or from about 60 to about 65% by weight water, or from about 65 to about 70% by weight water.

The microfibrillated cellulose obtainable by the present disclosure may comprise other optional additives including, but not limited to, dispersant, biocide, suspending aids, salt(s) and other additives, for example, starch or carboxy methyl cellulose or polymers, which may facilitate the interaction of mineral particles and fibres.

In certain embodiments in which a grindable inorganic particulate is present, the fibrous substrate comprising cellulose and inorganic particulate material are present in the aqueous environment at an initial solids content of at least about 2 wt %, of which at least about 2% by weight is fibrous substrate comprising cellulose, for example, an initial solids content of from about 2% by weight to about 20% by weight, or from about 4% by weight to about 15% by weight, or from about 5% by weight to about 12% by weight, or from about 7% by weight to about 10% by weight. In such embodiments, at least about 5% by weight of the initial solids content may be fibrous substrate comprising cellulose, for example, at least about 10%, or at least about 15%, or at least about 20% by weight of the initial solids content may be fibrous substrate comprising cellulose. In certain embodiments, no more than about 40% by weight of the initial solids content is fibrous substrate comprising cellulose, for example, no more than about 30% by weight of the initial solids content is fibrous substrate comprising cellulose, or no more than about 25% by weight of the initial solids content is fibrous substrate comprising cellulose

The grinding process may include a pre-grinding step in which coarse inorganic particulate is ground in a grinder vessel to a predetermined particle size distribution, after which fibrous material comprising cellulose is combined with the pre-ground inorganic particulate material and the grinding continued in the same or different grinding vessel until the desired level of microfibrillation has been obtained.

As the suspension of material to be ground may be of a relatively high viscosity, a suitable dispersing agent may be added to the suspension prior to or during grinding. The dispersing agent may be, for example, a water soluble condensed phosphate, polysilicic acid or a salt thereof, or a polyelectrolyte, for example a water soluble salt of a poly(acrylic acid) or of a poly(methacrylic acid) having a number average molecular weight not greater than 80,000. The amount of the dispersing agent used would generally be in the range of from 0.1 to 2.0% by weight, based on the weight of the dry inorganic particulate solid material. The suspension may suitably be ground at a temperature in the range of from 4° C. to 100° C.

Other additives which may be included during the microfibrillation step include: carboxymethyl cellulose, amphoteric carboxymethyl cellulose, oxidising agents, 2,2,6,6-Tetramethylpiperidine-1-oxyl (TEMPO), TEMPO derivatives, and wood degrading enzymes.

In certain embodiments, the product of the process is treated to remove at least a portion or substantially all of the water to form a partially dried or essentially completely dried product. For example, at least about 10% by volume, for example, at least about 20% by volume, or at least about 30% by volume, or least about 40% by volume, or at least about 50% by volume, or at least about 60% by volume, or at least about 70% by volume or at least about 80% by volume or at least about 90% by volume, or at least about 100% by volume of water in product of the grinding process may be removed. Any suitable technique can be used to remove water from the product including, for example, by gravity or vacuum-assisted drainage, with or without pressing, or by evaporation, or by filtration, or by a combination of these techniques. The partially dried or essentially completely dried product will comprise microfibrillated cellulose and optionally inorganic particulate material and any other optional additives that may have been added prior to drying. The partially dried or essentially completely dried product may be stored or packaged for sale. The partially dried or essentially completely dried product may be optionally re-hydrated and incorporated in papermaking compositions and other paper products, as described herein.

The fibrous substrate comprising cellulose may be microfibrillated in the presence of an inorganic particulate material to obtain microfibrillated cellulose having a d50 ranging from about 5 to μm about 500 μm, as measured by laser light scattering. The fibrous substrate comprising cellulose may be microfibrillated in the presence of an inorganic particulate material to obtain microfibrillated cellulose having a d50 of equal to or less than about 400 μm, for example equal to or less than about 300 μm, or equal to or less than about 200 μm, or equal to or less than about 150 μm, or equal to or less than about 125 μm, or equal to or less than about 100 μm, or equal to or less than about 90 μm, or equal to or less than about 80 μm, or equal to or less than about 70 μm, or equal to or less than about 60 μm, or equal to or less than about 50 μm, or equal to or less than about 40 μm, or equal to or less than about 30 μm, or equal to or less than about 20 μm, or equal to or less than about 10 μm.

The fibrous substrate comprising cellulose may be microfibrillated in the presence of an inorganic particulate material to obtain microfibrillated cellulose having a modal fibre particle size ranging from about 0.1-500 μm and a modal inorganic particulate material particle size ranging from 0.25-20 μm. The fibrous substrate comprising cellulose may be microfibrillated in the presence of an inorganic particulate material to obtain microfibrillated cellulose having a modal fibre particle size of at least about 0.5 μm, for example at least about 10 μm, or at least about 50 μm, or at least about 100 μm, or at least about 150 μm, or at least about 200 μm, or at least about 300 μm, or at least about 400 μm.

The fibrous substrate comprising cellulose may be microfibrillated in the presence of an inorganic particulate material to obtain microfibrillated cellulose having a fibre steepness equal to or greater than about 10, as measured by Malvern. Fibre steepness (i.e., the steepness of the particle size distribution of the fibres) is determined by the following formula:


Steepness=100×(d30/d70)

The microfibrillated cellulose may have a fibre steepness equal to or less than about 100. The microfibrillated cellulose may have a fibre steepness equal to or less than about 75, or equal to or less than about 50, or equal to or less than about 40, or equal to or less than about 30. The microfibrillated cellulose may have a fibre steepness from about 20 to about 50, or from about 25 to about 40, or from about 25 to about 35, or from about 30 to about 40.

A suitable procedure for characterising the particles size distribution of microfibrillated cellulose, and mixtures of inorganic particulate material and microfibrillated cellulose, is described in WO-A-2010/131016, at page 40, line 32 to page 41 line, 34.

The particulate ceramic grinding medium may have (i) a surface roughness of at least about 0.02 μm, or (ii) a mean coefficient of friction of at least about 0.10, or both (i) and (ii). The grinding medium is formed by sintering a composition comprising at least one of zirconia (ZrO2), e.g., ceria-stabilised zirconia, and alumina (Al2O3).

In certain embodiments, the composition comprises zirconia (ZrO2), meaning that the particulate ceramic grinding medium formed by sintering such a composition will contain a zirconia phase.

In certain embodiments, the composition further comprises from about 5 wt. % to about 25 wt. % ceria (Ce2O3), based on the total weight of the composition, for example, from about 10 wt. % to about 20 wt. % ceria, or from about 12 wt. % to about 18 wt. % ceria, or from about 10 wt. % to about 15 wt. % ceria, or from about 11 wt. % to about 14 wt. % ceria, or from about 11 wt. % to about 13 wt. % ceria. Additionally, the composition may comprise at least about 40 wt. % zirconia, for example, from about 40 wt. % to about 90 wt. % zirconia, or from about 40 wt. % to about 80 wt. % zirconia, or from about 50 wt. % to about 70 wt. % zirconia, or from about 55 wt. % to about 70 wt. % zirconia, or from about 60 wt. % to about 75 wt. % zirconia, or from about 65 wt. % to about 75 wt. % zirconia, or from about 65 wt. % to about 70 wt. % zirconia, based on the total weight of the composition. Additionally, the composition may comprise up to about 40 wt. % alumina, for example, up to about 30 wt. % alumina, or from about 1 wt. % to about 40 wt. % alumina, or from about 5 wt. % to about 30 wt. % alumina, or from about 10 wt. % to about 25 wt. % alumina, or from about 10 wt. % to about 20 wt. % alumina, or from about 12 to about 20 wt. % alumina, or from about 14 wt. % to about 20 wt. % alumina, or from about 14 to about 18 wt. % alumina.

In embodiments in which the composition comprises ceria and zirconia, or ceria, zirconia and alumina, the ceria and zirconia may be in the form of a ceria-stabilised zirconia. In certain embodiments, the ceria-stabilized zirconia comprises from about 10 wt. % to about 20 wt. % ceria, and up to about 90 wt. % zirconia, based on the total weight of the ceria stabilized zirconia, for example, from about 12 to about 18 wt. % ceria and up to about 88 wt. % zirconia, or from about 14 wt. % to about 16 wt. % and up to about 86 wt. % zirconia, or up to about 85 wt. % zirconia, or up to about 84 wt. % zirconia.

In certain embodiments, the ceria-stabilised zirconia comprises no more than about 2 wt. % iron oxide, for example, no more than about 1 wt. % iron oxide, or no more than about 0.75 wt. % iron oxide, or no more than about 0.5 wt. % iron oxide, or from about 0.1 wt. % to about 0.75 wt. % iron oxide, or from about 0.2 wt. % to about 0.6 wt. % iron oxide.

In certain embodiments, the composition comprises at least about 10 wt. % alumina with the balance ceria-stabilised zirconia (which may comprise a minor amount of iron oxide, as described above) in which the ceria-stabilized zirconia contains relative amounts of ceria and zirconia as described above. In certain embodiments, the composition comprises from about 10 wt. % to about 30 wt. % alumina, with the balance ceria stabilized zirconia, for example about 15 wt. % to about 25 wt. % alumina, with the balance ceria-stabilized zirconia.

In certain embodiments, the composition comprises from about 15 wt. % to about 25 wt. % alumina, from about 10 wt. % to about 15 wt. % ceria, and from about 50 wt. % to about 75 wt. % zirconia.

In certain embodiments, the particulate ceramic grinding medium is formed by sintering a composition comprising at least about 90 wt. % alumina, for example, at least about 95 wt. % alumina, or at least about 99 wt. % alumina, or at least about 99.5 wt. % alumina, or at least about 99.9 wt. %, or substantially 100 wt. % alumina. For example, the particulate grinding medium may be made by sintering an alumina-containing material, such as, for example, technical grade alumina, bauxite or any other suitable combination of oxides thereof.

In certain embodiments, the particulate ceramic grinding medium is obtainable by a method comprising:

    • a. obtaining, providing or making a composition comprising raw materials suitable for making the ceramic grinding medium;
    • b. mixing the composition comprising raw materials, forming a mixture;
    • c. combining the mixture with binder and/or solvent, forming a bound mixture;
    • d. granulating the bound mixture by mixing the bound mixture over a period of time during which the mixing speed is reduced;
    • e. optionally drying the granulated composition;
    • f. optionally shaping the granulated composition;
    • g. optionally sizing the granulated composition; and
    • h. sintering the granulated composition.

In certain embodiments, the raw materials in step b) of the method are homogenized, e.g., by mixing, forming a homogenized composition. By ‘homogenized’ is meant that the mixture of raw materials has a uniform composition throughout. In such embodiments, the homogenized composition is combined with binder and/or solvent in step c), forming a bound homogenized composition, which is granulated in step d) by mixing the bound homogenized composition over a period of time during which the mixing speed is reduced.

The binding agent and/or solvent is one of those well known in the industry. Possible binding agents include, for example, methyl cellulose, polyvinyl butyrals, emulsified acrylates, polyvinyl alcohols, polyvinyl pyrrolidones, polyacrylics, starch, silicon binders, polyacrylates, silicates, polyethylene imine, lignosulfonates, alginates, etc. In certain embodiments, a polyvinyl alcohol binder is used.

Possible solvents may include, for example, water, alcohols, ketones, aromatic compounds, hydrocarbons, etc.

Other additives well known in the industry may be added as well. For example, lubricants may be added, such as ammonium stearates, wax emulsions, oleic acid, Manhattan fish oil, stearic acid, wax, palmitic acid, linoleic acid, myristic acid, and lauric acid. Plasticizers may also be used, including polyethylene glycol, octyl phthalates, and ethylene glycol.

In certain embodiments, homogenizing comprises mixing the composition comprising raw materials for a suitable period of time such that the mixture of raw materials has a uniform composition throughout. In certain embodiments, step c) comprises mixing the homogenized composition with the binder and/or solvent. In certain embodiments, the mixing speed during step b) is greater than the mixing step in step c), and an initial mixing speed in step d) is no greater than a final mixing speed in step c).

In certain embodiments, mixing or homogenizing in step b) comprises mixing the composition comprising raw materials for a period of time from about 1 minute to about 60 minutes, for example, from about 1 minute to about 30 minutes, or from about 1 minute to about 20 minutes, or from about 1 minute to about 10 minutes, or from about 2 minutes to about 10 minutes, or from about 2 minutes to about 8 minutes, or from about 2 minutes to about minutes. Typically, the mixing speed is held constant during step b).

In certain embodiments, combining, e.g. mixing, the mixture or homogenized composition with the binder and/or solvent may be carried over a period of time of from about 30 seconds to about 30 minutes, for example, from about 30 seconds to about 20 minutes, or from about 30 seconds to about 10 minutes, or from about 1 minute to about 8 minutes, or from about 1 minute to about 5 minutes, or from about 2 minutes to about 5 minutes, or from about 2 minutes to about 4 minutes. As described above, the mixing speed during step c) is preferably less than the mixing speed in step b), and optionally at least the same as or greater than the initial mixing speed in step d). The binder and/or solvent may be added slowly during this step, e.g., continuously, or intermittently, preferably continuously. Alternatively, the all of the binder and/or solvent may be added at the beginning of mixing.

In certain embodiments, granulating the homogenized, bound composition, comprises mixing the composition over a period of time during which the mixing speed is gradually or stepwise reduced. A suitable period of time may be from about 1 minute to about 60 minutes, for example, from about 2 minutes to about 30 minutes, or from about 3 minutes to about 20 minutes, or from about 4 minutes to about 15 minutes, or from about 4 minutes to about 12 minutes, or from about 4 minutes to about 10 minutes, or from about 4 minutes to about 8 minutes. During the suitable period of time, the mixing speed may be reduced, e.g., stepwise, such that the final mixing speed is at least about 25% less than the initial mixing speed in step d), for example, at least about 30% less, or at least about 35% less, or at least about 40% less, or at least about 45% less than the initial mixing speed in step d).

In certain embodiments, an initial mixing speed in step b) is at least about 150% greater than a final mixing speed in step d), for example, at least about 175% greater, or at least about 190% greater, or at least about 200% greater, or at least about 210% greater.

The various mixing stages may be performed in any suitable mixing apparatus, for example, a mixer equipped with an impeller. An exemplary mixing apparatus is an Eirich mixer type RV02E equipped with a pin type impellor.

In certain embodiments, the initial impeller speed in step b) is between about 2750 and 3250 rpm, and the final impeller speed in step d) is between about 600 and 1200 rpm. In certain embodiments, the impeller speed in step b) is between about 2750 and 3250 rpm, and the impeller speed during step c) is between about 2000 and 2500 rpm. In such embodiments, the initial impeller speed in step d) is no greater than, preferably less than the impeller speed during step c), for example, less than about 2000 rpm, or less than about 1900 rpm, or less than about 1800 rpm. In such embodiments, the final impellor speed in step d) may be less than about 1500 rpm, for example, less than about 1200 rpm, or less than about 1000 rpm, or less than about 800 rpm. The final mixing speed, e.g., final impeller speed, may be held constant for a period of time ranging from about 1 minute to about 10 minutes, for example, from about 1 minute to about 8 minutes.

Following granulation, the granulated composition may be removed from the mixer and dried. For example, at a temperature of up to about 120° C. for a suitable period of time, e.g., from about 10 minutes to about 5 hours, or from about 30 minutes to about 2 hours. Before or during drying the granulated composition may be shaped, e.g., to form rod-shaped particles.

The optionally dried composition may then be subjected to a sizing process, e.g., by sieving. An appropriately sized sieve may be selected corresponding to the desired size of particulate grinding medium.

The particulated composition is then sintered at a suitable sintering temperature. Suitable sintering temperatures range from about 1200° C. to about 1700° C. The well time during sintering may range from about 1 hour to about 24 hours, for example, from about 2 hours to about 12 hours, or from about 2 hours to about 8 hours, or from about 2 hours to about 6 hours, or from about 3 hours to about 5 hours, or from about 3.5 hours to about 4.5 hours.

For embodiments in which the particulate ceramic grinding media is formed from a composition comprising at least ceria and zirconia, the sintering temperature is advantageously from about 1400° C. to about 1500° C., for example, from about 1425° C. to about 1475° C., or from about 1440° C. to about 1460° C., and a dwell time of from about 2 hours to about 6 hours, for example, from about 3 hours to about 5 hours, or from about 3.5 hours to about 4.5 hours.

For embodiments in which the particulated composition is formed from a composition comprising at least about 90 wt. % alumina, the sintering temperature is advantageously from about 1500° C. to about 1700° C., for example, from about 1550° C. to about 1650° C., or from about 1575° C. to about 1625° C., and a dwell time of from about 2 hours to about 6 hours, for example, from about 3 hours to about 5 hours, or from about 3.5 hours to about 4.5 hours.

In certain embodiments, the particulate grinding medium has a specific gravity of at least about 3.5, for example, a specific gravity of from about 3.5 to about 8.0, for example, from about 3.5 to about 7.0, or from about 3.5 to about 6.5, or a specific gravity of at least about 3.6, or at least about 3.7, or at least about 3.8, or at least about 3.9, or at least about 4.0, or at least about 4.1, or at least about 4.2, or at least about 4.3, or at least about 4.4, or at least about 4.5, or at least about 4.6, or at least about 4.7, or at least about 4.8, or at least about 4.9, or at least about 5.0, or at least about 5.1, or at least about 5.2, or at least about 5.3, or at least about 5.4, or at least about 5.5, or at least about 5.6, or at least about 5.6, or at least about 5.7, or at least about 5.8, or least about 5.9, or at least about 6.0.

In certain embodiments, the particulate grinding medium is used in the manufacture of microfibrillated cellulose. In certain embodiments, particulate grinding medium is used for improving one or more properties of the microfibrillated cellulose and/or for reducing the energy input per unit amount of microfibrillated cellulose produced.

In certain embodiments, the particulate grinding medium is used in a method for manufacturing microfibrillated cellulose, said method comprising a step of microfibrillating a fibrous substrate comprising cellulose by grinding in the presence of the particulate grinding medium which is to be removed after the completion of grinding.

According to certain embodiments, provided is an unpolished particulate grinding media having a surface roughness which increases by at least about 1% when subject to abrasive contact. By “unpolished” is meant that the grinding media has not been subjected to any polishing treatment (i.e., to smoothen its surface) prior to its use as a grinding media.

The increase in surface roughness may be determined in accordance with the methods described herein. In certain embodiments, the unpolished particulate grinding media has a surface roughness of at least about 0.5 μm, and/or (ii) a mean coefficient of friction of at least about 0.10 prior to abrasive contact. Abrasive contact may be an autogenous process (e.g., agitation in a mill or other suitable grinding apparatus) or may be conducted in the presence of another material, for example, another grinding media which, following abrasive contact, is separable from the unpolished particulate grinding media or, for example, a fibrous substrate comprising cellulose which, during abrasive contact, may be ground producing microfibrillated cellulose (e.g., microfibrillated cellulose according to embodiments described herein).

In certain embodiments, the surface roughness increases by at least about 5%, or at least about 10%, or at least about 15%, or at least about 20%, or at least about 25%, or at least about 30%, or at least about 35%, or at least about 40%, or at least about 45%, or at least about 50%. In certain embodiments, the material has a specific gravity of at least about 3.5. In certain embodiments, the unpolished particulate grinding media, prior to abrasive contact, has a surface roughness of at least about 2.0 μm, and/or (ii) a mean coefficient of friction of at least about 0.20, for example, a surface roughness of at least about 2.2 μm, or a surface roughness of at least about 2.4 μm, or a surface roughness of at least about 2.6 μm, or a surface roughness of at least about 2.8 μm, or a surface roughness of at least about 3.0 μm.

In certain embodiments, provided is a polished particulate grinding media having a surface roughness which increases by at least about 20% when subject to abrasive contact. By “polished” is meant that the grinding media has been subjected to a polishing treatment (i.e., to smoothen its surface) prior to its use as a grinding media. The increase in surface roughness may be determined in accordance with the methods described herein. In certain embodiments, the polished particulate grinding media has a surface roughness of at least about 0.5 μm, and/or (ii) a mean coefficient of friction of at least about 0.10 prior to abrasive contact. Abrasive contact may be an autogenous process (e.g., agitation in a mill or other suitable grinding apparatus) or may be conducted in the presence of another material, for example, another grinding media which, following abrasive contact, is separable from the polished particulate grinding media or, for example, a fibrous substrate comprising cellulose which, during abrasive contact, may be ground producing microfibrillated cellulose (e.g., microfibrillated cellulose according to embodiments described herein). In certain embodiments, the surface roughness increases by at least about 25%, or at least about 30%, or at least about 35%, or at least about 40%, or at least about 45%, or at least about 50%. In certain embodiments, the material has a specific gravity of at least about 3.5. In certain embodiments, the polished particulate grinding media, prior to abrasive contact, has a surface roughness of at least about 1.4 μm, and/or (ii) a mean coefficient of friction of at least about 0.08, or at least about 0.10, for example, a surface roughness of at least about 1.6 μm, or a surface roughness of at least about 1.8 μm, or a surface roughness of at least about 1.9 μm.

Method of Making Particulate Grinding Medium In certain embodiments, the particulate grinding medium may be made by any suitable method in which a particulate grinding having (i) a surface roughness of at least about 0.5 μm, or (ii) a mean coefficient of friction of at least about 0.10, or both (i) and (ii), is produced.

The method may comprise forming a particulate grinding medium which has a surface roughness of less than 0.5 μm and/or a mean coefficient of friction less than 0.10, and subjecting the particulate grinding medium to a surface roughening step such that the surface roughness is at least about 0.5 μm, and/or the mean coefficient of friction is at least about 0.10, at the end of the surface roughening step. For example, a particulate grinding medium initially not meeting the surface roughness and/or mean coefficient of friction requirements of the present disclosure may be co-ground with an abrasive material, such as a micro abrasive powder (e.g., a fused alumina micro abrasive powder, in a grinding vessel, such as a planetary mill.

Advantageously, the particulate grinding medium may be made by a process comprising:

    • a. obtaining, providing or making a composition comprising raw materials suitable for making the ceramic grinding medium;
    • b. mixing the composition comprising raw materials, forming a mixture;
    • c. combining the mixture with binder, forming a bound mixture;
    • d. granulating the bound mixture by mixing the bound mixture over a period of time during which the mixing speed is reduced;
    • e. optionally drying the granulated composition;
    • f. optionally shaping the granulated composition;
    • g. optionally sizing the granulated composition; and
    • h. sintering the granulated composition.

In certain embodiments, the raw materials in step b) of the method are homogenized, e.g., by mixing, forming a homogenized composition. In such embodiments, the homogenized composition is combined with binder and/or solvent in step c), forming a bound homogenized composition, which is granulated in step d) by mixing the bound homogenized composition over a period of time during which the mixing speed is reduced.

Paper Products and Processes for Preparing Same

The composition obtainable by the present invention comprising microfibrillated cellulose and (when present) inorganic particulate material can be incorporated in papermaking compositions, which in turn can be used to prepare paper products. The term paper product, as used in connection with certain embodiments of the present invention, should be understood to mean all forms of paper, including board such as, for example, white-lined board and linerboard, cardboard, paperboard, coated board, and the like. There are numerous types of paper, coated or uncoated, which may be made according to certain embodiments of the present invention, including paper suitable for books, magazines, newspapers and the like, and office papers. The paper may be calendered or super calendered as appropriate; for example super calendered magazine paper for rotogravure and offset printing may be made according to the present methods. Paper suitable for light weight coating (LWC), medium weight coating (MWC) or machine finished pigmentisation (MFP) may also be made according to the present methods. Coated paper and board having barrier properties suitable for food packaging and the like may also be made according to the present methods.

In a typical papermaking process, a cellulose-containing pulp is prepared by any suitable chemical or mechanical treatment, or combination thereof, which are well known in the art. The pulp may be derived from any suitable source such as wood, grasses (e.g., sugarcane, bamboo) or rags (e.g., textile waste, cotton, hemp or flax). The pulp may be bleached in accordance with processes which are well known to those skilled in the art and those processes suitable for use in certain embodiments of the present invention will be readily evident. The bleached cellulose pulp may be beaten, refined, or both, to a predetermined freeness (reported in the art as Canadian standard freeness (CSF) in cm3). A suitable paper stock is then prepared from the bleached and beaten pulp.

The papermaking composition typically comprises, in addition to the composition comprising microfibrillated cellulose and (when present) inorganic particulate material, paper stock and other conventional additives known in the art. For example, a papermaking composition may comprise up to about 50% by weight inorganic particulate material derived from the composition comprising microfibrillated cellulose and inorganic particulate material based on the total dry contents of the papermaking composition. For example, the papermaking composition may comprise at least about 2% by weight, or at least about 5% by weight, or at least about 10% by weight, or at least about 15% by weight, or at least about 20% by weight, or at least about 25% by weight, or at least about 30% by weight, or at least about 35% by weight, or at least about 40% by weight, or at least about 45% by weight, or at least about 50% by weight, or at least about 60% by weight, or at least about 70% by weight, or at least about 80% by weight, or at least about 90% by weight of inorganic particulate material derived from the composition comprising microfibrillated cellulose and inorganic particulate material, based on the total dry contents of the papermaking composition. The papermaking composition may also contain a non-ionic, cationic or an anionic retention aid or microparticle retention system in an amount in the range from about 0.1 to 2% by weight, based on the dry weight of the aqueous suspension comprising microfibrillated cellulose and inorganic particulate material. It may also contain a sizing agent which may be, for example, a long chain alkylketene dimer, a wax emulsion or a succinic acid derivative. The composition may also contain dye and/or an optical brightening agent. The composition may also comprise dry and wet strength aids such as, for example, starch or epichlorhydrin copolymers.

Paper products according to certain embodiments of the present invention may be made by a process comprising: (i) obtaining or preparing a fibrous substrate comprising cellulose in the form of a pulp suitable for making a paper product; (ii) preparing a papermaking composition from the pulp in step (i), the composition of certain embodiments of this invention comprising microfibrillated cellulose and (when present) inorganic particulate material, and other optional additives (such as, for example, a retention aid, and other additives such as those described above); and (iii) forming a paper product from said papermaking composition. As noted above, the step of forming a pulp may take place in the grinder vessel by addition of the fibrous substrate comprising cellulose in a dry state, for example, in the form of a dry paper broke or waste, directly to the grinder vessel. The aqueous environment in the grinder vessel will then facilitate the formation of a pulp.

An additional filler component (i.e., a filler component other than the inorganic particulate material which may be co-ground with the fibrous substrate comprising cellulose) can be added to the papermaking composition prepared in step (ii). Exemplary filler components are PCC, GCC, kaolin, or mixtures thereof. Paper products made from such papermaking compositions may exhibit greater strength (e.g., improved burst strength) compared to paper products comprising microfibrillated cellulose made by a comparable process in which the particulate grinding medium used in the process has at the beginning of grinding (i) a surface roughness which is less rough and/or (ii) a lesser mean coefficient of friction than that described herein. Similarly, paper products prepared from a papermaking composition according to certain embodiments of the present invention comprising inorganic particulate may exhibit a strength which is comparable to paper products comprising less inorganic particulate material. In other words, paper products can be prepared from a paper making composition according to certain embodiments of the present invention at higher filler loadings without loss of strength.

The steps in the formation of a final paper product from a papermaking composition are conventional and well know in the art and generally comprise the formation of paper sheets having a targeted basis weight, depending on the type of paper being made.

FLT Index

As used herein, “FLT Index” is a tensile strength measurement performed in accordance with the procedures of Example 2.

In some embodiments, the FLT index for the suspension of microfibrillated cellulose and the suspension of the microfibrillated cellulose and inorganic particulate material may reach 7.0 Nm/g or more, 7.5 Nm/g or more, 8 Nm/g or more, 8.5 Nm/g or more, 9 Nm/g or more, 9.5 Nm/g or more, 10 Nm/g or more, 10.5 Nm/g or more, 11 Nm/g or more, 11.5 Nm/g or more, 12 Nm/g or more, 12.5 Nm/g or more, 13 Nm/g or more, 13.5 Nm/g or more, 14 Nm/g or more, 14.5 Nm/g or more, or 15 Nm/g or more, through the methods disclosed in the specification.

In some embodiments, the FLT index may be improved by about 5%, about 6%, about 7%, about 7.5%, about 8%, about 9%, about 10%, about 12.5%, about 15% or about 20%, or more, compared to suspensions of microfibrillated cellulose or microfibrillated cellulose and inorganic particulate material produced without the application of ultrasonic energy treatment to the slurry or suspension.

The FLT index is a tensile test developed to assess the quality of microfibrillated cellulose and re-dispersed microfibrillated cellulose. The POP of the test material is adjusted to 20% by adding whichever inorganic particulate was used in the production of the microfibrillated cellulose/inorganic material composite (in the case of inorganic particulate free microfibrillated cellulose then 60 wt. %<2 μm GCC calcium carbonate is used). A 220 gsm sheet is formed from this material using a bespoke Buchner filtration apparatus. The resultant sheet is conditioned and its tensile strength measured using an industry standard tensile tester.

Matching Media Size to Fibre Type

Appropriate media size can be applied based upon the fibre species and the target FLT. In general a higher FLT can be achieved with a finer media, but costs more energy. The strength and length of the fibre are determinant of what media size to use.

Large media is good at fibre breakage but poor at fibrillation. Fibre breakage can be calculated as:

Fibre Breakage Factor ( mm - 1 ) = ( 1 / x p ) - ( 1 / x f )

    • where xp is the length-weighted fibre length (Lc (I)) of the MFC product (in mm) and xf is the length-weighted fibre length (Lc (I)) of the cellulosic fibre feed (in mm).

Fibre breakage must happen first for a given particle before effective fibrillation can begin. Therefore, the total fibrillation can be limited either by the fibrillation step or the fibre breakage step. To maximise fibrillation, the smallest media size possible should be used to give sufficient fibre breakage to permit a good rate of the subsequent fibrillation step, and this depends on the resistance of the fibre to breakage, which is thought to be dependent on the parameters in Table 1 below.

TABLE 1 Feed fibre properties. Lc(l) Fibre Coarseness Dry Zero-Span Tensile Fibre Length (mm) (mg/m) Index (Nm/g) NBSK (Botnia) 1.78 0.191 128 Eucalyptus 0.746 0.051 123 (UPM) Birch (Sodra) 0.887 0.08 139 Acacia (April) 0.735 0.042 93

The zero-span tensile strength is the strength of a sheet of fibres when the tensile test clamps are touching; therefore, each individual fibre is clasped with both clamps, and thereby the sheet fails by cross-sectional breakage of the fibres rather than network failure between the fibres (as is usually the case in long-span tensile testing). Zero-span tensile index is a measure of the fibre strength and is an indication of its resistance to breakage. Of the above, all three examples have a high zero-span strength, apart from acacia which is significantly lower.

NBSK is a long fibre with thick cross-section and has a high zero-span strength. It cannot be sensibly processed with fine media (e.g., 1.7 mm) without excessive energy input due to this. To discharge at relatively low energies necessitates large media (e.g., 5 mm). Sensible cascade grinding of NBSK includes a large media stage.

Eucalyptus and Birch fibres, on the other hand, although they are strong, are also thin and short. The thinness means less cross-section that needs breaking in a single collision, and the shortness means lower viscosity in a grinder and the fibre length does not need to be reduced as much before it can be discharged from the screen. Here, either fine media (e.g., 3 mm or 1.7 mm) can be practically used, with the one selected determined by whether a lower FLT at low energy input (e.g., 3 mm media) is desired or higher FLT at higher energy input (e.g., 1.7 mm media) is desired. Usually, 1.7 mm media is better, but there is often an intersection energy below which 3 mm media is better and allows for a significantly lower minimum discharge energy.

For acacia, in addition to being short and thin, they also have weaker cross-sections (i.e., low zero span strength); the additional collision energy of 3 mm media is unnecessary, and so 1.7 mm media is a better choice as it can still efficiently break the fibres down whilst also improving fibrillation.

Effect of Mineral Vs. Fibre Type and Media Size

In a similar principle to selecting the media size based upon fibre properties, as with a coarser media, the addition of mineral improves fibre breakage rate but harms fibrillation. Consequently, when fibre breakage is limiting and Lc (I) is not being reduced effectively, mineral addition can greatly improve this and so improve FLT. When fibre breakage is already reasonable at 100 POP, mineral addition in the grinder tends to be detrimental to FLT because it is not needed to enhance fibre breakage and instead its detrimental effects to liberated product fibrils is apparent.

Cascade Grinding

Cascade Grinding is the concept of using several grinders connected in series, and can be used to: (i) lower the energy input required to achieve a certain product quality; (ii) reduce the average product size; and (iii) improve the product quality achievable compared to single stage grinding in continuous mode alone. It does this by two mechanisms: (1) narrowing the residence time distribution; and (2) allowing for optimisation to maximise fibre breakage in early stages and maximise fibrillation in later stages.

The first mechanism is present for all forms of cascade grinding, even where conditions such as media size are not changed between stages (e.g. 3 mm/3 mm cascade). The second mechanism is present only when grinding conditions are changed between stages (e.g. 3 mm/1.7 mm cascade).

Cascade grinding is widely used for stirred mill processing of minerals for similar reasons as above; narrow residence time distribution and to optimise particle breakage at different stages in the comminution process.

Due to the generally superior FLT/energy relationship, a coarser media size can be opted for in the first stage and a finer media size in the second stage, with the exception of cases where oversize requirements are particularly strict, or discharge of product from the second stage with fine media is too difficult (in which case coarse media could be used in both stages).

Long, strong fibres, e.g. NBSK, benefit from a coarse media first stage (e.g. 5 mm) to give sufficient fibre breakage and sufficiently low discharge energy to make this practical. For short, strong fibres (e.g., eucalyptus and acacia), 3 mm media is sufficient as the coarse stage. In all cases tested, a fine media (1.7 mm media) second stage tended to give the best FLT/energy balance, with same-stage cascading with coarse media only being significantly beneficial for strictest oversize requirements or due to discharge problems in the second stage.

For example, a 5 mm/1.7 mm cascade is beneficial for NBSK, a 3 mm/1.7 mm cascade is beneficial for eucalyptus and birch.

Generally, cascade grinding allows for the production of high-strength products at reasonable energy inputs whilst also keeping the particle size low.

Mineral Free/Media Roughness

One of the barriers to creating a mineral-free MFC product using stirred media detritors for applications where the presence of mineral is deleterious to performance (e.g. in barrier or packaging applications) is the presence of a substantial amount of worn grinding media particles contaminating the product. Understanding what factors influence this wear and how to produce it is key to producing a viable product for mineral-free applications. Media wear is deleterious in many of these applications as (i) as with other mineral particles, it can add weight and degrade mechanical and barrier properties without providing useful benefits, (ii) higher media wear results in greater grinding media consumption, increasing the production costs.

Since the presence of micron-scale wear particles in the product is largely a consequence of attrition wear rather than compressive strength failure, it was hypothesised that the surface roughness of the grinding media was a key factor in controlling wear rate, with the higher roughness media having asperities more at risk of being broken off.

There are two sets of experiments discussed below. The first demonstrates the difference in media wear between two commercially available mullite medias with different roughnesses. The second is a broader study to demonstrate that the difference seen in the first experiment is in fact due to the roughness and not something else, and that there is an optimum roughness for the right balance of energy efficiency and product contamination from media wear.

EXAMPLES Example 1. Interferometer Operation Omniscan MicroXAM2

1. Switch power on.

2. Boot up PC.

3. Adhere a grinding media bead to a glass slide.

4. Locate the specimen particle directly under the light beam, preferably focusing directly on top of the particle. An image will appear on the screen/monitor, which will not be clear (blurry).

5. Alter the light intensity so that there is a red spot in the middle of the picture (the red spot should not cover the full image on the screen).

6. Check if the red spot becomes smaller on moving the lens down (anti clockwise on dial) towards the particle. The image becoming more out of focus.

7. Then bring the lens back up to the position it was before, and then turn the light intensity down so that the red dot is much smaller and less defined.

8. Then slowly keep moving the lens up (clockwise on dial) until the image comes into focus (the particle surface becomes more defined). Turn the light intensity down if needed so the red light is more sparse and less bold.

9. When the image is in focus, tare the position of the lens on the control box.

10. Then run the sample (after entering the correct file name).

11. With the image that is displayed you can crop out any anomalies by using the crop button on the left, and right clicking to select ‘make main image’. Then save.

12. Open the saved image in the ProfilmOnline software associated with the equipment.

13. Using the ‘line roughness’ tool, record the output arithmetic roughness (Ra) of the particle at least 5 times in the horizontal direction, and at least 5 times in the vertical direction, ensuring that representative sections of the sample are taken.

14. Calculate the mean Ra value of these 10+ line samples.

15. Repeat with additional media beads to a sufficient manner to be confident that the sample is representative of the batch.

16. Note that for particularly low-roughness and for particularly fine media (high curvature), often towards the boundaries of the image, the signal reflected to the detector is weak due to insufficient surface exposed at the appropriate angle. This can generate artefacts of sharp, narrow roughness spikes, indicating higher roughness values than the true values (near the centre-point of the media). These artefacts are obvious when the measured roughness at the centre-point (top) of the media bead (where angle of reflection is closer to incident beam) is substantially lower than at the borders.

17. For situations such as the above where significant artefacts are present, either the image should be cropped to exclude the vast majority of these border artefacts, or the lines drawn in the software to calculate roughness are retracted away from the border to exclude the border areas where there are artefacts. In these manners, to better ensure a representative measurement of the surface morphology, care must be taken to ensure that as much of surface is included in this measurement as possible, only excluding the artefact-heavy regions.

Example 2. FLT Test

The FLT test is a quick measurement for making sheets from a pure MFC sample on a custom-built filtration apparatus, which is utilized to measure the strength of a MFC sheet.

Table 2 below shows the calibration data from mineral-free NSBK ground at 2% fibre solids. FIG. 2 shows the comparison of these calibrations. Similar tensile strength procedures are known in the art, such as TAPPI T-404 cm-92, which is incorporated herein by reference in its entirety. The FLT test performed in the present Examples followed the procedure noted below.

Performance of FLT Tensile Strength Measurement.

Apparatus utilized was a Tensile test filtration apparatus, 18.5 cm diameter medium-fast speed filter papers (Whatman No. 40 or equivalent).

The % solids and Percentage of Pulp (% POP) of sample were recorded as determined in the separate Examples described below.

Record % solids and % POP of sample (see separate procedures described below).

If % POP is greater than 20%, add GCC with a particle size of 60%<2 μm (e.g. Imerys IC60 (IC60)) (see separate procedure for MFC handsheets). If % POP is between 18% and 20% a correction factor will need to be applied to the result.

Approximately 4.4 g dry weight of sample (44 g for a 10% solids sample) was taken and diluted with water to 400 mL to obtain a total solids of approximately 1.1% (0.22% fibre solids). This will make a 220 gsm sheet on the 15.9 cm diameter exposed screen of the apparatus. The sample was stirred well to ensure good dispersion.

Then, 1 ml-3 ml depending on the expected FLT) of the 0.2 wt % polyDADMAC solution was added to the diluted sample and the sample was stirred well. The top section of the filtration unit was removed and a filter paper was placed on top of the screen.

Thereafter, the filter paper was wetted with a wash bottle, and any bubbles that formed were pushed out to the rim of the paper. The vacuum was then switched on to adhere the filter to the screen, ensuring that it sat flush with no creases. The top section of the apparatus was clamped in place and the vacuum was switched off and the drain valve was opened to release vacuum and drain water. The sample was poured into the top section over the end of a spatula or similar instrument to ensure an even distribution. Pouring the sample directly onto filter paper was avoided. The sample was allowed to settle for a few seconds, then the vacuum was switched on and the sample was filtered. This took approximately 2 minutes. Once the water cleared, the vacuum supply was switched off after approximately 1 minute and the drain valve was opened to release the vacuum and to remove water from unit. Thereafter, the top section of unit was removed and the filter paper and filtered sample together were carefully removed. The sample and filter were carefully placed on a Rapid Kothen carrier board. The sheet cover of the Rapid Kothen was placed over the sample and dried in the Rapid Kothen drier for approximately 7 minutes. The dry sample was separated from the filter and cover and conditioned at 23° C. and 50% RH±5% RH for a minimum of 20 minutes Next, the sheet was weighed to determine its grams per square meter (“gsm”). The sample was cut into 15 mm wide strips using a cutter. A minimum of 5 strips were required.

Next, the force in Newtons required to break each strip with the tensile tester was measured.

Calculations of FLT were made in the following manner.

Area of sheet in m 2 ( A ) = 0.001 × Π × ( diameter in cm ) 2 / 4 ( 0.0199 for 15.9 cm diamter sheet ) . Sheet g s m = Mass of sheet in grams / A . Mass of slurry required = 100 × 200 × A / TS ( TS = % total solids ) . F L T cm kg - 1 ( T ) = 1000 × Fm / ( W × g s m ) .

Where Fm=Max tensile force (N).

W=Strip width (15 mm as standard).

gsm=gsm of sample

The average tensile index and standard deviation of the 5 measurements in each case were recorded.

As noted above, if the % POP is less than 20%, then the tensile index is corrected according to:

Tcorrected = T / [ 1 - 7 .6 * ( 0.2 - % POP ) ] .

Calibration and procedures follow those laid down in Paper testing-T220 sp-96.

TABLE 2 Summary of production conditions and testing results for the calibration grinds for NBSK, ground at 2% fibre solids. Malvern Insitec Specific 25- 150- Sample energy Total <25 150 300 >300 Vane FLT and/or input solids POP D30 D50 D70 D90 μm μm μm μm Viscosity index Conditions kWh/t % % μm μm μm μm Steepness % % % % mPas N · m/g X274 9.2 1.7.19 (Control) Calibration 2500 0.7 99.4 116 242 531 1151 22 5.6 31.3 18.9 44.3 Too Dilute 6.0 curve Lab 3000 0.8 99.8 117 241 516 1139 23 5.7 31.0 19.3 44.0 Too Dilute 6.6 Grind - 3500 1.2 99.5 126 260 507 1044 25 5.7 28.8 19.8 45.8 3240 8.4 Min-free 4000 0.8 99.8 123 251 526 1143 23 5.6 29.7 19.7 45.0 Too Dilute 6.8 Sodra 4500 1.0 99.3 120 251 517 1124 23 6.1 29.5 19.4 44.9 2980 6.4 Blue @ 5000 1.3 99.4 108 221 424 844 26 7.0 31.7 20.6 40.7 3120 8.6 2% Fibre 5500 1.6 99.3 74 149 272 551 27 10.3 39.9 23.0 26.8 2520 10.5 Solids

FLT test is performed at 20 dry wt. % MFC; in this instance the samples were diluted using additional host mineral (IC60) to 20 dry wt. % MFC so that it could be compared to experimental control samples produced at 20 dry wt. % MFC.

Example 3. Lab Scale Grinds

TABLE 3 Input parameters used for lab-scale grinds in Example 3 Parameter Values Grinder Used Laboratory-scale 500 W Cheetah Stirred Media Mill Grinder Mode Batch Fibre Species NBSK Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) For 3 mm: 0.20 μm; for (for each media size) 1.7 mm: 0.10 μm POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 Media Volume Concentration (% MVC)  42 Specific Energy Input (kWh/t) 1500; 3000; 4500

NBSK Lab Scale: Input Parameters in Table 3

FIGS. 1a and 1b shows plots from a lab grind of NBSK. FIG. 1a is a plot of Lc (I) vs energy input at 100 POP. FIG. 1b is a plot of FLT vs energy input at 100 POP. Since the Lc (I) reduction rate is very low for 1.7 mm media, fibrillation is breakage-limited and so increasing to a larger media size with more kinetic energy greatly increases the Lc (I) reduction (fibre breakage rate) and consequently leads to a large improvement in FLT. In general, when 3 mm media greatly reduces Lc (I) compared to 1.7 mm media, 3 mm media will give better FLT. When 3 mm media only slightly reduces Lc (I) compared to 1.7 mm media, 1.7 mm media will give better FLT. At some intermediate difference in Lc (I) reduction, 3 mm media is better at low energy inputs and 1.7 mm media is better at high energy inputs.

Example 4

NBSK pilot scale: (input parameters in Table 4) Pilot-scale 50 POP (IC60 GCC) batch grinds with NBSK were performed to compare 4 different media sizes. As can be seen in FIG. 2, the preferred media size depends on the target FLT. Coarse media (5 mm) gives the highest FLT at low energy inputs, but has a low peak FLT; 3 mm media gives a higher peak FLT but requires more energy, and 1.7 mm media is the same but to a greater extent. Excessively large media (10 mm) is poorer at all energy inputs.

TABLE 4 Input parameters used for pilot-scale grinds in Example 4. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Batch Fibre Species NBSK Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 10 mm; 5 mm; 3 mm; 1.7 mm Media Roughness (Ra) For 10 mm: 0.58 μm; for 5 mm: 0.30 μm; (for each media size) for 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm POP 100% Fibre Solids For 10 mm, 5 mm; 3 mm media: 2%; for 1.7 mm media: 1% Power Density (kW/m3) For 5 mm, 3 mm, 1.7 mm media: 65; for 10 mm media: 50 Media Volume For 10 mm, 5 mm, 1.7 mm media: 48; for Concentration (% MVC) 3 mm media: 42 Specific Energy 500-5000 Input (kWh/t)

NBSK full scale: Tables 6 and 7 below summarise the FLT and energy data for both 100 POP and 50 POP (GCC IC60). As can be seen, unlike with other shorter fibres, 1.7 mm media, although it gives higher FLT and finer Lc (I) results, requires an excessive amount of energy input (unless cascade grinding is used).

TABLE 5 Input parameters used for full-scale grinds in Example 4. Parameter Values Grinder Used Full-scale 355 kW stirred media mill Grinder Mode Continuous Fibre Species NBSK Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 5 mm; 3 mm; 1.7 mm Media Roughness (Ra) For 5 mm: 0.12 μm; for 3 mm: 0.16 μm; (for each media size) for 1.7 mm: 0.10 μm POP 100%; 50% Fibre Solids For 5 mm and 3 mm: 1.5%; for 1.7 mm: 1% Power Density 65 (kW/m3) Media Volume 42.5 Concentration (% MVC) Specific Energy 1250-5900 kWh/t Input (kWh/t)

TABLE 6 NBSK full scale continuous mode data at 100 POP. Specific Energy Input FLT Lc(l) Series (kWh/t) (Nm/g) (mm) Single Stage 5 mm media 1250 5.8 0.615 Single Stage 3 mm media 3000 9.7 0.512 Single Stage 1.7 mm media 5900 12.6 0.419

TABLE 7 NBSK full scale continuous mode data at 50 POP. Specific Energy Input FLT Lc(l) Series (kWh/t) (Nm/g) (mm) Single Stage 5 mm media 1500 5.5 0.544 Single Stage 3 mm media 2950 9.5 0.46 Single Stage 1.7 mm media 5000 10.7 0.372

Example 5. Matching Media Size to Fibre Type-Eucalyptus

Eucalyptus lab scale: (Inputs in Table 8) FIG. 3a shows the length-weighted fibre length Lc (I) of MFC produced by processing bleached Eucalyptus fibres with 3 mm and 1.7 mm mullite media, and FIG. 3b shows the FLT of these grinds. Here, from FIG. 3b the FLT achieved with 1.7 mm media is similar or slightly higher than that for 3 mm media. FIG. 3a shows that the length reduction rate of the 3 mm media is considerably greater than for 1.7 mm media, though the difference is significantly less than that seen with NBSK fibres (FIG. 1a). Consequently, negative effect of the reduction in breakage rate on FLT 10 from using finer media is cancelled out from an improvement in the fibrillation ability that the finer media provides.

TABLE 8 Input parameters used for lab-scale grinds in Example 5. Parameter Values Grinder Used Laboratory-scale 500 W Cheetah Stirred Media Mill Grinder Mode Batch Fibre Species Eucalyptus Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) For 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 Media Volume 42 Concentration (% MVC) Specific Energy 1500; 3000; 4500 Input (kWh/t)

Eucalyptus pilot scale: (inputs in Table 9) FIG. 4 provides a comparison of 3 mm and 1.7 mm media continuous grinds for Eucalyptus at pilot scale. 1.7 mm media ultimately gave the highest strength, but 3 mm media had a lower discharge energy. The pilot scale data gives essentially the same conclusions as the full scale data discussed below.

TABLE 9 Input parameters used for pilot-scale grinds in Example 5. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Continuous Fibre Species Eucalyptus Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) for 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100% Fibre Solids  1% Power Density 50 (kW/m3) Media Volume For 3 mm media: 48; for 1.7 mm media: 42. Concentration (% MVC) Specific Energy 500-5500 Input (kWh/t)

Eucalyptus full scale: (Inputs in Table 10) Table 11 provides data for a full scale continuous grind of eucalyptus at 50 POP (IC60 GCC) using both 3 mm and 1.7 mm media, whereas Table 12 provides data for a full scale continuous grinds of eucalyptus at 100 POP using both 3 mm and 1.7 mm media. FIGS. 5 and 6 provide data for a full scale continuous grind of eucalyptus at 100 POP. FIG. 5 shows that there is an intersection point between 1.7 mm and 3 mm media curves, above which 1.7 mm media is better and below which 3 mm media is better. FIG. 6 shows that although lower FLT results are achieved with 3 mm media, it gives products with finer MFC fibre lengths. Tables 5 and 6 shows that, for Eucalyptus, a greater FLT can be obtained using 1.7 mm media when energy input is held constant, and that this is true both whether it is co-ground with mineral (Table 11) or in the absence of mineral (Table 12). Table 13 shows that, for Eucalyptus, a FLT of 9 can be obtained with less energy input when using 1.7 mm media.

TABLE 10 Input parameters used for full-scale grinds in Example 5. Parameter Values Grinder Used Full-scale 355 kW stirred media mill Grinder Mode Continuous Fibre Species Eucalyptus Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) 3 mm: 0.16 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100%; 50% (mineral: IC60 GCC) Fibre Solids For 3 mm: 1.5%; for 1.7 mm: 1.25% Power Density (kW/m3) 65 Media Volume 42.5 Concentration (% MVC) Specific Energy Input (kWh/t) 1250-3000

TABLE 11 Eucalyptus full scale continuous data for 50 POP. Specific Energy FLT Series Input (kWh/t) (Nm/g) Single Stage 3 mm media 2500 8.2 Single Stage 1.7 mm media 2500 9.9

TABLE 12 Eucalyptus full scale continuous data for 100 POP. Specific Energy FLT Series Input (kWh/t) (Nm/g) Single Stage 3 mm media 3000 9.9 Single Stage 1.7 mm media 3000 11.2

TABLE 13 Eucalyptus full scale continuous data for 100 POP. Energy Needed for FLT Series 9 Nm/g Single Stage 3 mm media 2500 Single Stage 1.7 mm media 2000

Example 6. Matching Media Size to Fibre Type-Birch

Birch pilot scale: (inputs in Table 14) FIG. 7 provides data from a pilot plant continuous grinding of, birch at 100 POP, comparing 3 mm and 1.7 mm media. FIG. 7 shows similar conclusions as with the eucalyptus data (although it is noted that FIG. 7 does not go to sufficiently high energy with the 3 mm curve to see the benefit of 1.7 mm media at higher energy).

TABLE 14 Input parameters used for pilot-scale grinds in Example 6. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Continuous Fibre Species Birch Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) for 3 mm: 0.16 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100% Fibre Solids  1% Power Density For 3 mm media: 65; for 1.7 mm media: 50 (kW/m3) Media Volume 42 Concentration (% MVC) Specific Energy 1000-5500 Input (kWh/t)

Birch full scale: (inputs in Table 15) As shown in Table 16 below, for birch a greater FLT can be obtained using 1.7 mm media when energy input is held constant.

TABLE 15 Input parameters used for full-scale grinds in Example 6. Parameter Values Grinder Used Full-scale 355 kW stirred media mill Grinder Mode Continuous Fibre Species Birch Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) 3 mm: 0.16 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100% Fibre Solids  1% Power Density (kW/m3) 50 Media Volume 42.5 Concentration (% MVC) Specific Energy Input (kWh/t) 2750

TABLE 16 Birch full scale continuous data for 100 POP. Specific Energy Input Series (kWh/t) FLT (Nm/g) Single Stage 3 mm media 2750 12.1 Single Stage 1.7 mm media 2750 13

Example 7. Matching Media Size to Fibre Type-Acacia

Acacia lab scale: (inputs in Table 17) FIGS. 8a and 8b show data from batch lab grinds with 3 mm and 1.7 mm media for acacia fibres. FIG. 8a shows Lc (I) vs. energy input. FIG. 8b shows FLT vs. energy input. There is little difference in the breakage rate between the media sizes; 1.7 mm media has sufficient energy to break the fibres down, making 3 mm media redundant. Consequently, the benefit of the better fibrillation of fine media is maximised, and the resultant FLT with 1.7 mm media is much greater than that of 3 mm media. This effect is much stronger than for Eucalyptus fibres (FIGS. 3a and 3b), which since fibre dimensions are otherwise similar (see Table 1), can be attributed to the lower zero-span tensile index.

TABLE 17 Input parameters used for lab-scale grinds in Example 7. Parameter Values Grinder Used Laboratory-scale 500 W Cheetah Stirred Media Mill Grinder Mode Batch Fibre Species Acacia Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) For 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm (for each media size) POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 Media Volume 42 Concentration (% MVC) Specific Energy 1500; 3000; 4500 Input (kWh/t)

Acacia pilot scale: (inputs in Table 18) FIG. 9 shows pilot-scale batch acacia grinds at 50 POP (IC60 GCC). FIG. 9 shows there is no benefit of 3 mm media even at low energy inputs.

TABLE 18 Input parameters used for pilot-scale grinds in Example 7. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Batch Fibre Species Acacia Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm; 1 mm Media Roughness (Ra) for 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm; (for each media size) for 1 mm: 0.082 μm POP 50% (Mineral: IC60 GCC) Fibre Solids 1.5% Power Density 50 (kW/m3) Media Volume 42 Concentration (% MVC) Specific Energy 1000-5500 Input (kWh/t)

Acacia full scale: (inputs in Table 19) Table 20 shows that slightly lower energy input can be used while obtaining a slightly higher FLT with 1.7 mm media than with 3 mm media.

TABLE 19 Input parameters used for full-scale grinds in Example 7. Parameter Values Grinder Used Full-scale 355 kW stirred media mill Grinder Mode Continuous Fibre Species Acacia Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) 3 mm: 0.20 μm; for 1.7 mm: 0.10 μm (for each media size) POP 50% (mineral: IC60 GCC) Fibre Solids For 3 mm: 1.5%; for 1.7 mm: 1.25% Power Density (kW/m3) 65 Media Volume 42.5 Concentration (% MVC) Specific Energy Input (kWh/t) 2250-2500

TABLE 20 Full scale continuous mode grind of acacia at 50 POP. Specific Energy FLT Series Input (kWh/t) (Nm/g) Single Stage 3 mm media 2500 5.3 Single Stage 1.7 mm media 2250 6.5

Example 8. Matching Media Size to Fibre Type-Various Fiber Types

Cross-fibre data: (inputs in Tables 5, 10, 15, and 19) Table 21 includes full-scale data at 100 POP showing the properties of each fibre, and the FLT and energy inputs achieved with each media size trialed. NBSK, as a long, strong fibre requires excessive energy inputs with fine media, and requires less fine (e.g., 3 mm) media. For Eucalyptus and Birch, which are short and strong, fine media can be utilised at moderate energy inputs while still leading to an improvement in product quality.

Table 22 includes full-scale data at 50 POP (IC60 GCC) showing the properties of each fibre, and the FLT and energy inputs achieved with each media size trialed. NBSK, as a long, strong fibre gives excessive energy inputs with fine media, and requires less fine (e.g., 3 mm) media. For Eucalyptus and Acacia, which are short, fine media can be utilised at moderate energy inputs while still leaving to an improvement in product quality.

TABLE 21 Full-scale data at 100 POP showing the properties of each fibre, and the FLT and energy inputs achieved with each media size trialled. MFC FLT/Energy vs. Media Size (Full Scale) 100 POP Continuous Fibre Properties NBSK Eucalyptus Birch Fibre Length Lc(l) (mm) 1.78 0.746 0.887 Fibre Coarseness (mg/m) 0.191 0.051 0.08 Fibre Dry Zero Span 128 123 139 Tensile Index (Nm/g) MFC Properties Ground with Energy FLT Energy FLT Energy FLT Media Size: (kWh/t) (Nm/g) (kWh/t) (Nm/g) (kWh/t) (Nm/g) 5 mm mullite 1250 5.8 3 mm mullite 3000 9.7 3000 9.9 2750 12.1 1.7 mm mullite 5900 12.6 3000 11.2 2750 13.0

TABLE 22 Full-scale data at 50 POP showing the properties of each fibre, and the FLT and energy inputs achieved with each media size trialled. MFC FLT/Energy vs. Media Size (Full Scale) 50 POP Continuous Fibre Properties NBSK Eucalyptus Acacia Fibre Length Lc(l) (mm) 1.78 0.746 0.735 Fibre Coarseness (mg/m) 0.191 0.051 0.042 Fibre Dry Zero Span 128 123 93 Tensile Index (Nm/g) MFC Properties Ground with Energy FLT Energy FLT Energy FLT Media Size: (kWh/t) (Nm/g) (kWh/t) (Nm/g) (kWh/t) (Nm/g) 5 mm mullite 1500 5.5 3 mm mullite 2950 9.5 2500 8.2 2500 5.3 1.7 mm mullite 5000 10.7 2500 9.9 2250 6.5

TABLE 23 Input parameters used for lab-scale grinds in Example 8. Parameter Values Grinder Used Laboratory-scale 500 W Cheetah Stirred Media Mill Grinder Mode Batch Fibre Species NBSK, Black Spruce, Southern Pine, Dissolving Pulp (NBSK-based), Birch, Eucalyptus, Acacia, Mixed European hardwood, Tissue dust (hardwood fibres from dust extraction in tissue mill) Cotton linters, Abaca, Miscanthus Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 2.9 mm Media Roughness (Ra) 0.20 μm POP 50% Fibre Solids 2.5%  Impeller Speed (RPM)  800 Media Volume Concentration  48 (% MVC) Specific Energy Input (kWh/t) 3000

Lab-scale grinds with various fibre species: (inputs in Table 23).

FIG. 10 is a plot of MFC tensile index for various fibres ground with 1 and 2.9 mm media, produced at 50% POP (IC60 GCC) in lab scale batch mode at an energy input of 3000 kWh/t. Here, long, strong fibres such as softwoods (e.g. Nordic Pine) and certain stronger non-woods (Abaca, cotton) produce relatively poor MFC when processed with 1 mm media, with the coarser 2.9 mm media being superior. The opposite is true for hardwood fibres, weak softwoods (dissolving pulp) and other non-woods (e.g. miscanthus), where 1 mm media leads to a better quality MFC than 2.9 mm media.

FIG. 11 is a plot of the ratio of MFC tensile index between MFC produced with fine media and coarse media versus fibre zero-span tensile index, produced at 50% POP (IC60 GCC) in lab scale batch mode at an energy input of 3000 kWh/t. Here, the influence of fibre type and fibre zero-span tensile index is demonstrated; for short fibres (hardwoods and short non-woods), the ratio is above 1, so the finer media produces a better product. However, decreasing fibre zero-span tensile index increases this ratio, increasing the benefit of using finer media. For longer fibres (softwoods and long non-woods), the ratio is largely below 1, meaning that coarser media is superior. However, at sufficiently low fibre zero-span tensile index values, the ratio increases above 1, resulting in the finer media being superior.

FIG. 12 is a plot of Lc (I)-based operating index versus feed fibre length for 2.9 and 1 mm media grinds, produced at 50% POP (IC60 GCC) in lab scale batch mode at an energy input of 3000 kWh/t. KR,Op is the Operating Rittinger Index, and is an internally-derived measure of the resistance of a particle to breakage. The higher this is, the harder it is to reduce Lc (I). Here, for 1 mm media, a rapid increase in KR,Op is seen with increasing fibre length, demonstrating that long fibre lengths greatly decrease the effectiveness of fibre breakage for finer media. For 2.9 mm media, KR,Op does not show an obvious general correlation with fibre length, so coarser media are less subject to restrictions in fibre breakage efficiency due to longer fibre lengths.

FIG. 13 is a plot of MFC Lc (I) values for various fibre species ground with 1 and 2.9 mm media, produced at 50% POP (IC60 GCC) in lab scale batch mode at an energy input of 3000 kWh/t. Here, it is clear that in all cases, 2.9 mm media greatly decreases fibre length, whereas 1 mm media is always less efficient at doing so. In some cases, 1 mm media is clearly incompetent at breaking fibres, for example for softwoods and abaca.

FIGS. 14a-h are differential interference contrast microscopy images of MFC produced by 1 mm and 2.9 mm media grinds of various fibre species, produced at 50% POP (IC60 GCC) in lab scale batch mode at an energy input of 3000 kWh/t. FIGS. 14a-d show the results of 1 mm media grinds of Nordic pine, eucalyptus, mixed European hardwood, and tissue dust, respectively. FIGS. 14e-h show the results of a 2.9 mm media grinds of Nordic pine, eucalyptus, mixed European hardwood, and tissue dust, respectively. In all cases, the 2.9 mm media competently disintegrated the fibres into MFC. However, for weak fibres (mixed European hardwood (g) and tissue dust (h)) the fibrils produced were very short. Decreasing the media size for these fibres to 1 mm) and (d) respectively) produced longer liberated fibrils whilst still competently disintegrating the feed fibres, which FIG. 10 shows leads to a higher quality product. For Nordic pine, long fibrils were produced with 2.9 mm media (e) and reducing the media size to 1 mm resulted in a large fraction of the fibres remaining relatively unbroken (a).

Example 9. Effect of Mineral vs. Fibre Type and Media Size

NBSK lab scale: The 100 POP grinds in Table 4 were compared to 50 POP grinds (mineral: IC60 GCC) produced under otherwise identical conditions. FIG. 15a is a plot of Lc (I) vs. energy input for batch lab grinds with NBSK comparing 3 mm and 1.7 mm media. FIG. 15b is a plot of FLT vs. energy input for batch lab grinds with NBSK comparing 3 mm and 1.7 mm media. Considering 3 mm media, the Lc (I) reduction is improved significantly with mineral addition, which appears to counteract the detrimental effect of mineral on fibrillation and give a similar FLT-energy relationship as with 100 POP. The fibre breakage rate is even more limiting with 1.7 mm media due to lower collision kinetic energy, so mineral addition greatly enhances this, and so greatly improves tensile strength up to the level expected with 3 mm media.

NBSK full scale: In addition to the data in Table 5, an additional 20 POP (IC60 GCC) grind using 3 mm media under otherwise the same conditions were carried out to allow for a broader comparison of the influence of mineral content. FIG. 16 is a plot of full-scale continuous grinds with several media sizes comparing 100 POP to 50 and 20 POP. FIG. 17 is a plot of Lc (I) vs. energy input for full-scale continuous grinds with NBSK. Table 24 shows the effect of mineral addition on full scale continuous grinding of NBSK with 3 mm media. FIG. 17 shows that the benefit of mineral addition on size reduction rate is more obvious with decreasing media size.

As expected, with 3 mm media the energy-FLT relationship may not differ significantly with mineral addition, but the reduction in Lc (I) (apparent in FIG. 17) permits discharge at significantly lower energy with 20 POP. With 5 mm media, the mineral is not needed and 100 POP is better. With 1.7 mm media, although fibre breakage rate is helped with mineral addition, it is still too coarse to discharge from the grinder at sensible energy inputs.

TABLE 24 The effect of mineral addition on full scale continuous grinding of NBSK with 3 mm media. Energy at Tensile Fibre Mineral Discharge Index FLT Length Content (kWh/t) (Nm/g) Lc(I) (mm) 100 POP  3000 9.7 0.512 50 POP 2950 9.5 0.460 20 POP 2350 8.7 0.356

Eucalyptus full scale: Table 25 shows data for full scale continuous grinding of eucalyptus using the conditions described in Table 10. With 3 mm media, fibre breakage rate is already quite high, so mineral addition is not needed to help this, and instead it just degrades fibrillation; even at low energy inputs, mineral addition makes strength worse. With 1.7 mm media, this is true also at higher energy inputs, but at lower energy inputs where with 1.7 mm media there is an intersection point (at around 1850 kWh/t) below which it can be expected that the increased breakage rate from mineral addition is helpful.

TABLE 25 Data for full scale continuous grinding of eucalyptus. Specific Energy Input FLT (Nm/g) Media Size (kWh/t) 100 POP 50 POP 3 mm 1250 7.1 5.7 3 mm 2500 9.1 8.2 1.7 mm 1850 8.5 8.5 1.7 mm 3000 11.2 10.2

FIG. 18 is a plot of Lc (I) vs. energy input at full scale with eucalyptus in continuous mode, comparing 100 POP and 50 POP (IC60 GCC) grinds. Although mineral co-grinding increases the efficiency of fibre length reduction, in this case, as Table 25 shows, this leads to a degradation in FLT.

Acacia lab scale: Lab scale batch mode grinds were performed for acacia fibres, using conditions identified in Table 17, alongside 50 POP (IC60 GCC) series with both media sizes under otherwise identical conditions. It was observed that, for both 1.7 and 3 mm media sizes, mineral addition gives little benefit in the Lc (I) reduction rate (see FIG. 47). Therefore, it does not improve the rate of fibrillation, and the mineral degrades existing fibrils, leading to large drops in tensile strength (see FIG. 45).

Acacia pilot scale: (inputs in Table 26) FIG. 19 is a plot of pilot-scale continuous grinds with acacia fibres, comparing 100, 50, and 25 POP (IC60 GCC) grinds with 1.7 mm media.

TABLE 26 Input parameters used for pilot-scale grinds in Example 9. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Batch Fibre Species Acacia Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 1.7 mm Media Roughness (Ra) 0.10 μm POP 100%; 50%; 20% (Mineral: IC60 GCC) Fibre Solids 1.25% Power Density (kW/m3) 65 Media Volume Concentration (% MVC) 42.5 Specific Energy Input (kWh/t) 1000-3000

Across all energy inputs, 100 POP is better, since the presence of mineral here degrades fibrillation. This is also seen in Table 27.

TABLE 27 Acacia pilot scale continuous data. Specific Energy Input FLT (Nm/g) Media Size (kWh/t) 100 POP 50 POP 25 POP 1.7 mm 1500 6.5 6.2 6.1 1.7 mm 3000 9.2 8.0 7.0

General effect of mineral at lab scale: Where Lc (I) reduces the most from mineral addition, mineral is likely to be helpful (breakage limited). Where mineral-addition does not improve Lc (I) reduction much, mineral will be harmful (fibrillation limited). In general, the more the work index (measure of resistance to fibre breakage) goes down with mineral addition, the better the increase in FLT from mineral addition. If work index does not decrease by much, mineral addition is harmful to FLT.

Example 10. Cascade Grinding-NBSK

NBSK pilot scale: (inputs in Table 28) It is difficult to discharge pilot grinder with NBSK in continuous mode with 3 mm media; consequently, the energy input of the single stage control is very high, at 3750 kWh/t. Using 5 mm media as a first stage of cascade grinding helps size reduction rate, greatly reducing discharge energy and improving the FLT—energy relationship. Consequently, a better quality can be achieved at a lower energy input by cascade grinding. Several attempts were carried out at pilot scale; one was at 100 POP, with 4 stages, 5, 3, 1.7, and 1 mm media in sequence (see FIG. 20). A 50 POP (IC60 GCC) equivalent was done with 5, 3, and 1.7 mm in sequence (see FIG. 21), along with a 5 mm/1.7 mm two stage cascade for comparison. (see FIG. 22). A comparison of FIGS. 21 and 22 shows only a slight advantage of using a 3-stage process over a 2-stage process.

TABLE 28 Input parameters used for pilot-scale grinds in Example 10. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Continuous (Cascade) Fibre Species NBSK Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 5 mm; 3 mm; 1.7 mm; 1 mm Media Roughness (Ra) 5 mm: 0.30 μm; 3 mm: (for each media size) 0.20 μm; 1.7 mm; 0.10 μm; 1 mm: 0.082 μm POP 100%; 50% Fibre Solids 5 mm: 1.5%; 3 mm, 1.7 mm, 1 mm: 1% Power Density (kW/m3) 5 mm: 65, 3 mm, 1.7 mm, 1 mm: 50 Media Volume Concentration (% MVC) 5 mm, 3 mm: 48; 1.7 mm, 1 mm: 42 Specific Energy Input (kWh/t) 1000-3750

NBSK full scale: Using full-scale grinding conditions as identified in Table 5, the conventional single-stage continuous products were compared with cascade variants which use the same conditions for each media size as identified in Table 5, but with the energy divided between two grinders, with the first grinder containing a larger media size than the second grinder.

Although better than the single stage controls, the full scale attempts were not fully optimised and do not show a strong benefit compared to the standard 3 mm media single stage control. The exception is the 100 POP 5 mm/1.7 mm cascade, which gives a strong product for only somewhat higher energy. FIG. 23 is a plot of FLT vs. energy input for full-scale continuous data for NBSK, 100 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes). FIG. 24 is a plot of MFC fibre length Lc (I) vs. energy input for full-scale continuous data for NBSK, 100 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes). FIG. 25 is a plot of FLT vs. energy input for full-scale continuous data for NBSK, 50 POP, comparing single stage results (circles) with 2-stage cascade results (other shapes). FIG. 26 is a plot of MFC fibre length Lc (I) vs. energy input for full-scale continuous data for NBSK, 50 POP (IC60 GCC), comparing single stage results (circles) with 2-stage cascade results (other shapes). Table 29 shows FLT tensile strength, MFC fibre length Lc (I), and specific energy input at discharge, for NBSK at full scale and 100 POP, comparing cascade grinding with single stage series. Table 30 shows FLT tensile strength, MFC fibre length Lc (I), and specific energy input at discharge, for NBSK at full scale and 50 POP, comparing cascade grinding with single stage series.

TABLE 29 FLT tensile strength, MFC fibre length Lc(I), and specific energy input at discharge, for NBSK at full scale and 100 POP, comparing cascade grinding with single stage series. The energy inputs in parentheses are the stage energies for the 1st and 2nd stages respectively. Specific Energy FLT Lc(I) Series Input (kWh/t) (Nm/g) (mm) Single Stage 3 mm media 3000 9.7 0.512 Single Stage 1.7 mm media 5900 12.6 0.419 Cascade 5 mm/3 mm 3000 (1250/ 9.8 0.364 1750) Cascade 5 mm/1.7 mm 3650 (1300/ 12.8 0.348 2350)

TABLE 30 FLT tensile strength, MFC fibre length Lc(I), and specific energy input at discharge, for NBSK at full scale and 50 POP, comparing cascade grinding with single stage series. The energy inputs in parentheses are the stage energies for the 1st and 2nd stages respectively. Specific Energy FLT Lc(I) Series Input (kWh/t) (Nm/g) (mm) Single Stage 3 mm media 2950 9.5 0.46 Single Stage 1.7 mm media 5000 10.7 0.372 Cascade 5 mm/1.7 mm 3500 (1500/ 10.6 0.285 2000)

FIG. 23 and Table 29 demonstrate that the 5 mm/1.7 mm cascade at 100 POP gives a significantly stronger product than the 3 mm single stage control, at somewhat more energy input. To get a similar strength as that cascade series (>12 Nm/g) in single stage requires 1.7 mm media with an excessive cost (5900 kWh/t). By using cascade grinding, this high strength can be achieved with an energy saving of 2250 kWh/t. FIG. 24 and Table 29 shows that the cascade grinds give significantly finer products than either of the single stage controls, with the 5 mm/3 mm case giving the same energy input and FLT as the 3 mm single stage, but much finer size; this could be important for some applications. FIGS. 25-26 and Table 30 demonstrate the same principles as the above, but at 50 POP (IC60 GCC).

Example 11. Cascade Grinding-Eucalyptus

Eucalyptus pilot scale: (inputs in Table 31) FIG. 27 is a plot of FLT-energy results for pilot-scale continuous grinding with Eucalyptus (unless otherwise stated, using 100 POP and 42 MVC) comparing various routes for single stage grinding, two-stage cascade grinding. From this, it is clear that cascade grinding improves the FLT at a given energy input or lowers the energy input to achieve a certain FLT; this is particularly true when the second stage media is substantially finer than the first stage media. Additionally, the use of multiple stages decreases the product fibre length more rapidly, which is important for some size-sensitive applications.

TABLE 31 Input parameters used for pilot-scale grinds in Example 11. Cascade variants listed show the sequential medias in each grinder in series, which were compared against single-stage controls of each media size tested. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Continuous (Cascade) Fibre Species Eucalyptus Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) 3 mm: 0.16 μm; (for each media size) 1.7 mm; 0.10 μm Media Sizes/Combinations Trialled 3 mm alone; 1.7 mm alone; 3 mm/3 mm; 3 mm/1.7 mm; 3 mm/3 mm/1.7 mm. POP All 100%, except an additional 25 POP 3 mm/1.7 mm cascade Fibre Solids 1% Power Density (kW/m3) All grinds 65 kW/m3, except the 48 MVC variant which was 73 kW/m3 Media Volume Concentration (% MVC) All grinds 42 MVC, except an additional 48 MVC 3 mm/1.7 mm cascade Specific Energy Input (kWh/t) 750-6000 kWh/t. In all cases, when cascade is used, energy input is divided equally between each stage.

Eucalyptus full scale: Single stage full-scale grinds were carried out as in Table 10, and compared with two cascade arrangements under otherwise the same conditions as the single stage media for the relevant media size (with the exception of 3 mm media size, wherein the fibre solids in cascade mode was 1.25% versus 1.5% in single stage mode). Two cascade variants were trialled: one with substantially the same media size in the first and second stages (3 mm/3 mm), and one where the media size in the second stage was substantially smaller than the first stage (3 mm/1.7 mm). In both cases, total energy input was divided equally between both stages. FIG. 31 is a plot of FLT vs. energy for full-scale 100 POP eucalyptus continuous grinds, comparing single stage to cascade grinding. FIG. 32 is a plot of MFC fibre length Lc (I) vs. energy for full-scale 100 POP eucalyptus continuous grinds, comparing single stage to cascade grinding. Table 32 is shows results from a comparison of 100 POP eucalyptus fibres ground in continuous mode with different single stage and cascade arrangements. FIGS. 31 and 32 show that a minor improvement in FLT is achieved in single stage using cascade grinding versus the best single stage (1.7 mm media), but compared to that example, a much finer product size is achieved.

TABLE 32 Comparison of 100 POP eucalyptus fibres ground in continuous mode with different single stage and cascade arrangements. FLT achieved Lc(I) achieved by 3000 by 3000 Series kWh/t (Nm/g) kWh/t (mm) Single Stage 1.7 mm media 11.2 0.494 Single Stage 3 mm media 9.9 0.399 Cascade 3 mm/3 mm media 10.7 0.337 Cascade 3 mm/1.7 mm media 11.8 0.36

Example 12. Cascade Grinding-Birch

Birch pilot scale: (inputs in Table 33) there are similar conclusions for Birch as with Eucalyptus. FIG. 33 is a plot of FLT-energy results for pilot-scale continuous grinding with Birch at 100 POP. FIG. 34 is a plot of MFC fibre length Lc (I) vs. energy results for pilot-scale continuous grinding with Birch at 100 POP. Using FIGS. 33 and 34, Tables 34-36 show comparisons either of the FLT and Lc (I) achievable using a certain route, or (for Table 37) the energy required to achieve a high FLT value of 14 Nm/g. These data demonstrate that the cascade variants show efficiency improvements in achieving a target FLT, and FLT improvements at a given target energy input.

TABLE 33 Input parameters used for pilot-scale grinds in Example 12. Cascade variants listed show the sequential medias in each grinder in series, which were compared against single-stage controls of each media size tested. Parameter Values Grinder Used 700 Pilot Mill (22 kW stirred media mill) Grinder Mode Continuous (Cascade) Fibre Species Birch Media Composition Mullite/calcined kaolin-based Media Specific Gravity 2.8 g/cm3 Media Size (d50) 3 mm; 1.7 mm Media Roughness (Ra) 3 mm: 0.16 μm; (for each media size) 1.7 mm; 0.10 μm Media Sizes/Combinations Trialled 3 mm alone; 1.7 mm alone; 3 mm/3 mm; 3 mm/1.7 mm; 3 mm/3 mm/1.7 mm. POP 100 Fibre Solids 1% Power Density (kW/m3) 65 kW/m3 for 42 MVC, 73 kW/m3 for 48 MVC Media Volume Concentration (% MVC) 42 MVC and 48 MVC Specific Energy Input (kWh/t) 1000-6000 kWh/t. In all cases, when cascade is used, energy input is divided equally between each stage.

TABLE 34 Birch pilot grinds continuous 100 POP 2000 kWh/t. Birch Pilot Grinds Continuous 100 POP 2000 kWh/t Series FLT (Nm/g) Lc(I) (mm) Single stage 3 mm (42 MVC) 9.6 0.542 Cascade 3 mm/3 mm (42 MVC) 11.1 0.462 Cascade 3 mm/1.7 mm (48 MVC) 12.3 0.483

TABLE 35 Birch pilot grinds continuous 100 POP 3000 kWh/t. Birch Pilot Grinds Continuous 100 POP 3000 kWh/t Series FLT (Nm/g) Lc(I) (mm) Single stage 1.7 mm (42 MVC) 10 0.635 Cascade 3 mm/1.7 mm (42 MVC) 12.4 0.463 Cascade 3 mm/3 mm/1.7 mm (42 MVC) 14.2 0.404

TABLE 36 Birch pilot grinds continuous 100 POP 6000 kWh/t. Birch Pilot Grinds Continuous 100 POP 6000 kWh/t Series FLT (Nm/g) Lc(I) (mm) Single stage 1.7 mm (42 MVC) 14.1 0.515 Cascade 3 mm/1.7 mm (42 MVC) 16.3 0.336

TABLE 37 Birch pilot grinds continuous 100 POP. Birch Pilot Grinds Continuous 100 POP Energy required to reach an Series FLT of 14 Nm/g (kWh/t) Single stage (1.7 mm) 6000 Two stage cascade (3 mm/1.7 mm) 4000 Three stage cascade (3 mm/3 mm/1.7 mm) 3000

Example 13. Cascade Grinding for Lowering Energy Input for a Target FLT

FIG. 27 demonstrates that one can use cascade grinding to lower the energy input required to achieve a target FLT. For example, one example of cascade grinding can achieve FLT 9 Nm/g by 1500 kWh/t, which is half of the energy that a 3 mm media grind would require to get to this. Although the data is not there, this will be more efficient than 1.7 mm media if full-scale equivalent results are considered, although the efficiency improvement is less compared to 3 mm media. It is much easier to make the comparison of the FLT achievable with a given energy budget since the energy budget is a controlled factor.

Example 14. Cascade Grinding for Maximising Achievable FLT

For some applications which require the highest performance products, under-processed material which can be bypassed through the grinder in continuous mode may hinder product performance. Using cascade grinding, extensiveness of this bypassing can be reduced, and therefore improve the maximum strength achievable; with single stage 1.7 mm media, the FLT peaks at 11.6 Nm/g, whereas with a 3 mm/1.7 mm cascade, an FLT of 13.3 Nm/g is achievable. FIG. 28 is a plot of MFC fibre length Lc (I) versus energy results for pilot-scale continuous grinding with Eucalyptus (unless otherwise stated, using 100 POP and 42 MVC) comparing various routes for single stage grinding. FIG. 29 is a plot of FLT and Lc (I) achieved by an energy input of 1500 kWh/t, comparing single stage to different cascade grind variants. FIG. 30 is a plot of FLT and Lc (I) achieved by an energy input of 3000 kWh/t, comparing single stage to different cascade grind variants. Table 38 shows results of a pilot-comparison of 100 POP eucalyptus fibres ground in continuous mode with different single stage and cascade arrangements.

TABLE 38 Pilot-comparison of 100 POP eucalyptus fibres ground in continuous mode with different single stage and cascade arrangements. FLT Lc(l) FLT Lc(l) achieved achieved achieved achieved by 1500 by 1500 by 3000 by 3000 kWh/t kWh/t kWh/t kWh/t Series (Nm/g) (mm) (Nm/g) (mm) Single Stage 1.7 mm 10.1 0.511 media Single Stage 3 mm 6.5 0.491 9.1 0.422 media Cascade 3 mm/3 mm 7.7 0.451 9.8 0.34 media Cascade 3 mm/1.7 mm 9 0.472 11.5 0.381 media

Example 15. Cascade Grinding for Size Reduction Whilst Maintaining Quality

The data in Example 11 and Table 32 demonstrate that a 3 mm/3 mm cascade gives higher FLT than single stage 3 mm but also gives the lower Lc (I), producing a finer product; this is entirely due to the narrowing of the residence time distribution. Replacing the second stage of this cascade with a fine 1.7 mm stage leads to the highest FLT of the series tested, although somewhat reduces the effectiveness of Lc (I) reduction (although still better than either single stage arrangement tested). On its own, at 3000 kWh/t, the single stage 1.7 mm media grinds give MFC with a good tensile strength, but it is very coarse and would be less suitable for oversize-sensitive applications. By cascade grinding, we can achieve equivalent or better FLT than the best-performing single stage grind, but reduce the fibre length of the product considerably.

Example 16. Mineral Free/Media Roughness Data

Comparison of rougher and smoother 3 mm mullite media at 100 POP:

Surface roughness measurement was carried out as described in Example 1. Two formulations of 3 mm mullite were tested; one having a surface roughness of approximately 3 times higher than the second formulation. The smoother formulation had a surface roughness of 0.20 μm, and the coarser formulation had a surface roughness of 0.63. The difference in surface was visibly obvious from the white light interferometer images (see FIG. 35).

Lab grinds: Standard lab grinds were carried out with both grinding media species, (47.5% MVC, 800 rpm, 2.5% fibre solids), in the absence of added mineral, at specific energy inputs varying from 500 kWh/t to 5000 kWh/t. Since no mineral was added other than worn grinding media generated in the grinding, the level of grinding media contamination was assessed from how much the POP value deviated from 100% (Media On Fibre (MOF)=(100%-% POP)/% POP). The resultant MFC was tested for the FLT tensile strength, and tested with the Valmet FS5 fibre image, particularly to analyse changes in product fibre length.

Media wear results: FIG. 36 shows the dramatic reduction in grinding media wear when switching to the smoother formulation of grinding media. Since media wear is approximately linear with energy input, the relative media wear of each can be expressed using the gradients of these curves with energy (as % MOF contamination per unit energy input); this is shown in Table 39, and indicates that the wear rate of the rougher media is around seven times higher than that of the smoother media. The smoother media thereby produces a product that is much more suitable for use in applications where the presence of worn media is not desired.

TABLE 39 Gradient of the media wear vs. energy curve for each series. Rougher Smoother Mullite 100 Mullite 100 POP POP Gradient of % MOF vs. Energy 5.22 0.752 Curve (%/1000 kWh/t)

Fibre length results: Although there is a dramatic improvement in media wear when the smoother media is used, the effect of the lower roughness on the efficiency and effectiveness of the production process needed to be assessed. Observing the length-weighted fibre length of the MFC products using both media types (see FIG. 37) demonstrates that the rougher media is significantly more effective at cutting the fibrous particles into smaller units. This is an unexpected conclusion for stirred mill comminution, as in their typical use in the minerals processing industry, the media roughness is typically disregarded and not considered to be important; in fact, the limited data available that investigates this (Skuse, T., 2015, “The use of Positron Emission Particle Tracking (PEPT) to Determine the Grinding Mechanisms within a Vertically Stirred Media Mill,” PHD Thesis, Department of Chemical Engineering, University of Birmingham, United Kingdom) indicates that the surface roughness of the grinding media has no significant effect on the particle size reduction rate of mineral particles. The fact that for fibres the improvement that surface roughness of the grinding media brings to size reduction rate is therefore significant.

FLT tensile index results: The FLT tensile index of MFC produced with both media species is shown as FIG. 38. Here, it is obvious that the rougher media has significantly improved the tensile strength of the MFC compared to the smoother media. The enhancement of the fibre breakage rate as demonstrated by FIG. 37 also allows for more rapid development of fibrillation and therefore higher tensile strength.

Conclusions: Although the rougher media produces a higher quality product more efficiently, it comes at a cost of a much higher media wear rate that precludes its use in many applications. Therefore, it can be considered worthwhile that the selection of a smoother grinding media is worth the additional energy cost to achieve a certain product performance.

Example 17. General Effect of Roughness on Wear and Grinding Efficiency

Although the previous study gave circumstantial evidence that surface roughness of the grinding media largely controls wear rate and grinding efficiency, better proof was sought by obtaining a particular species of grinding media and modifying it to change only the roughness and no other physical property. In order to do this, some glass grinding media was acquired of the same media size (3 mm) and similar density as mullite (2.5 g/cm3 compared to 2.8 g/cm3). As received, glass grinding media is very smooth, with a Ra value as measured of 0.022 μm. Using this as a feedstock, various batches of this glass media were roughened using a stirred media detritor to grind a slurry of an abrasive mineral of various particle sizes and composition (either silicon carbide or mullite powder).

This provided a range of 12 glass grinding media batches with Ra roughness values varying from the 0.022 μm feed, up to 1.16 μm for the roughest batch. Grinds were carried out with each media batch using the same conditions as in Example 16, but to specific energy inputs of 1000 kWh/t and 3000 kWh/t only. This was supplemented with data from the previously used mullite media study, along with two mullite media species from different manufacturers that had intermediate roughnesses of Ra=0.40 and 0.41 μm.

Also included are alumina and zirconium silicate grinding medias of 3 mm in size, but of different roughness as purchased from the manufacturer, and zirconia media that was roughened in the same way as the glass media. This is not wholly a fair comparison of just roughness in these cases, as density of the grinding media also differs, but overall it demonstrates a more universal trend.

Effect of roughness on wear rate: FIGS. 39 and 40 show the effect of media roughness on the wear rate for each media composition when ground to an energy of 3000 kWh/t. The data from the roughened glass in particular demonstrates the strong correlation between media surface roughness and the wear rate. The mullite data also show an increase with media wear, though it is a less strong correlation due to some difference in composition between the samples (i.e., it is not the same batch of sample where the only difference changed is media roughness). FIG. 40 also shows a general increase with zirconia and alumina/zirconium silicate. Roughness is not the sole factor that determines media wear rate, material composition and compressive strength also do, and so the curves for the latter are more noisy due to differences in composition.

Effect of roughness on MFC fibre length: FIGS. 41 and 42 show the MFC length-weighted fibre length at 1000 kWh/t and 3000 kWh/t respectively, demonstrating an obvious decay in fibre length with increasing roughness. Additionally, despite probable difference in media wear rate with composition, the difference in Lc (I) between the different media compositions at a given surface roughness is relatively minor, indicating that roughness is a much more important factor than the other differences between the media (i.e., density). Notably, the glass and mullite grinding media (where density is very similar) follow the same curve, demonstrating that roughness is the determining factor for MFC size reduction.

Effect of roughness on MFC FLT tensile index: FIG. 43 shows the influence of media roughness on FLT at both 1000 kWh/t and 3000 kWh/t. At 1000 kWh/t, it is noted that a higher media roughness clearly improves the FLT. At 3000 kWh/t, it is noted that regardless of media species, FLT rapidly increases with roughness at low roughness values, before coming to a plateau and slightly declining at higher roughnesses. This demonstrates that there is an optimum roughness value above which there are diminishing returns in terms of FLT when increasing roughness further (where it is known that this would greatly accelerate the media wear rate). At 3000 kWh/t, this optimum roughness appears to be in the region of Ra=0.15-0.4 μm. However, since media wear increases with increasing roughness, it can be more preferable to operate at the lower end of this range, and even below it (i.e. down to around 0.1 μm), since further increases in roughness are not significantly improving FLT, yet are increasing contamination from worn media wear.

Conclusions: By selecting a media roughness in the region of Ra=0.15-0.4 μm, a moderately strong FLT can be achieved at moderate energy inputs, whilst maintaining media wear at reasonably low levels (on the order of ˜5% MOF). The selection of the 3 mm mullite media of Ra=0.20 appears to be a beneficial choice in terms of the balance between energy efficiency and product contamination from media wear. The data above suggests that if media roughness were reduced significantly lower than Ra=0.2 μm, although wear rate would improve, the energy required to achieve a moderately strong product would increase excessively under the processing conditions used herein (particularly obvious at Ra<0.05 μm in this example).

Example 18. Media Size Selection

TABLE 40 Lab-scale batch grind parameters. Roughness Band 0.25 μm-0.6 μm Scale Lab Grinder Mode Batch Fibre NBSK (Metsa Botnia Pine) POP 50% Fibre Solids  1% Impeller Speed (RPM) 600, 1000 RPM MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 3 mm, 1.7 mm, 1 mm Roughness estimated at about mullite 0.3 μm based upon measurements of other media

TABLE 41 Lab-scale batch grind parameters and results. Media Impeller Media Roughness Energy FLT Speed Size (Interferometer Input index Lc(l) (RPM) (mm) Ra) (μm) (kWh/t) (Nm/g) (mm) 1000 3 0.2 1500 6.97 0.699 1000 1.7 0.10 1500 5.18 1.282 1000 1 0.082 1500 5.95 1.485 600 3 0.2 1500 6.23 0.84 600 1.7 0.1 1500 4.46 1.529 600 1 0.082 1500 1.648 1000 3 0.2 3000 7.75 0.411 1000 1.7 0.1 3000 8.33 0.645 1000 1 0.082 3000 8.65 0.955 600 3 0.2 3000 8.99 0.408 600 1.7 0.1 3000 0.899 600 1 0.082 3000 6.36 1.119

This Example 18 relates to lab-scale batch grinds carried out using 3 mm, 1.7 mm, and 1 mm media, co-grinding with IC60 GCC at 50% POP and 1% fibre solids. Table 41 shows that for targets of lower FLT (the data at 1500 kWh/t energy input), the 3 mm media gives a product with significantly higher FLT, which appears to be because the finer media is unable to cause sufficient fibre breakage (Lc (I) reduction) to permit efficient fibrillation. At higher energy inputs, there has been more time for fibre breakage, so the finer media is improved disproportionately compared to the 3 mm media. The optimum media size to use depends partly on the target product quality, with larger media being efficient at low quality targets, whereas vice versa is the case for smaller media.

Example 19. Effect of Aggressive Grinding Conditions, as Increased by Media Size, Roughness, or Impeller Speed, on Fibre Breakage Rate and Fibrillation

TABLE 42 Lab-scale batch grind parameters. Media Roughnesses Tested 0.02 μm-0.8 μm Scale Lab Grinder Mode Batch Fibre NBSK (Metsa Botnia Pine) POP 100% Fibre Solids  2.5% Impeller Speed (RPM) 300, 400, 500, 800, 1050, 1100, 1200, 1500 MVC (%) 47.5 Energy Inputs (kWh/t) 3000 Many media sizes of different compositions and roughnesses tested

The data in this example are all lab-scale batch grinds using NBSK fibres at a fibre solids content of 2.5%, 100% POP, and an MVC of 48%, at a specific energy input of 3000 kWh/t. The media size, density, and roughness were changed by changing media species, and impeller speed was varied in order to change the kinetic energy of a collision between media of given properties. The media wear was measured (% MOF), along with the FLT and the Lc (I). This was in order to understand the effect of more aggressive grinding conditions, as increased by media size, roughness, or impeller speed, on the fibre breakage rate and fibrillation of the product. These data are displayed in Table 43 below.

In general, the level of fibre breakage, as measured by Lc (I) reduction, was greater in cases when (i) media roughness was higher, (ii) media size was greater, and (iii) impeller speed was greater. In general, such conditions lead to high FLT tensile indexes also, but it is also possible to use excessively intense conditions (excessively large media size, roughness, or impeller speed) that damages liberated fibril quality and reduces product tensile index.

A higher media roughness permits the use of less intense conditions of the other parameters (media size and impeller speed) whilst maintaining a high product quality, but this comes at a cost of a higher media wear rate. There is therefore generally a trade-off between media wear in the product and FLT tensile index obtained, see FIG. 44.

TABLE 43 Lab-scale batch grinds using media of various sizes, densities, and roughnesses, at various impeller speeds, and the resultant MFC Lc(l), FLT tensile index, and % MOF media wear. MFC MFC Length- Tensile weighted Media Media Media Impeller Media index Fibre Media Size Density Roughness Speed Wear (FLT) Length Material (mm) (g/cm3) Ra (μm) (RPM) (% MOF) (Nm/g) Lc(l) (mm) Glass 3.11 2.48 0.022 800 1.48 5.91 1.250 Glass 3.11 2.48 0.022 1500 12.32 7.24 0.449 Glass 3.11 2.48 0.022 400 0.48 2.83 1.756 Glass 3.11 2.48 0.022 1050 4.34 7.16 0.670 Glass 1.81 2.48 0.028 800 0.76 3.54 1.659 Ceria-zirconia 0.97 5.99 0.038 300 0.90 2.88 1.867 Ceria-zirconia 0.97 5.99 0.038 500 1.53 3.18 1.836 Ceria-zirconia 0.97 5.99 0.038 1100 0.90 5.61 1.501 Ceria-zirconia 0.97 5.99 0.038 1500 0.66 6.13 1.093 Ceria-zirconia 2.49 6.02 0.044 300 1.48 3.30 1.628 Ceria-zirconia 2.49 6.02 0.044 500 0.14 4.10 1.305 Ceria-zirconia 2.49 6.02 0.044 800 0.97 7.52 0.615 Ceria-zirconia 2.49 6.02 0.044 1100 0.40 8.16 0.403 Ceria-zirconia 2.49 6.02 0.044 1500 1.45 7.58 0.435 Ceria-zirconia 2.49 6.02 0.044 1500 1.45 7.58 0.435 Zirconium 2.87 4.06 0.050 800 1.52 7.95 0.645 silicate Ceria-zirconia 1.51 6.04 0.069 300 0.89 3.47 1.796 Ceria-zirconia 1.51 6.04 0.069 500 1.48 3.30 1.717 Ceria-zirconia 1.51 6.04 0.069 800 0.22 4.74 1.398 Ceria-zirconia 1.51 6.04 0.069 1100 0.63 5.76 1.078 Ceria-zirconia 1.51 6.04 0.069 1500 0.63 7.64 0.496 Ceria-zirconia 2.89 6.04 0.072 300 0.60 3.79 1.540 Ceria-zirconia 2.89 6.04 0.072 500 0.76 5.15 1.179 Ceria-zirconia 2.89 6.04 0.072 800 0.69 7.88 0.515 Ceria-zirconia 2.89 6.04 0.072 800 0.83 9.47 0.447 Ceria-zirconia 2.89 6.04 0.072 1100 1.24 8.98 0.414 Ceria-zirconia 2.89 6.04 0.072 1500 1.37 8.20 0.444 Ceria-zirconia 2.89 6.04 0.072 1500 1.37 8.20 0.444 Zirconium 2.89 4.05 0.10 1500 1.95 6.67 0.450 silicate Zirconium 2.89 4.05 0.10 400 0.30 3.78 1.534 silicate Zirconium 2.89 4.05 0.10 800 1.16 9.80 0.447 silicate Ceria-zirconia 2.11 6.04 0.11 300 0.81 4.02 1.681 Ceria-zirconia 2.11 6.04 0.11 500 1.09 4.66 1.358 Ceria-zirconia 2.11 6.04 0.11 800 0.48 8.02 0.539 Ceria-zirconia 2.11 6.04 0.11 1100 0.90 9.24 0.343 Ceria-zirconia 2.11 6.04 0.11 1500 0.74 7.57 0.368 Glass 3.11 2.48 0.16 800 3.90 9.77 0.467 Mullite 2.96 2.80 0.20 800 2.60 10.34 0.394 Yttria-zirconia 2.77 6.05 0.22 800 3.04 9.33 0.305 Yttria-zirconia 2.77 6.05 0.23 800 1.23 9.94 0.406 Glass 3.11 2.48 0.25 800 5.30 10.48 0.411 Mullite 1.66 2.69 0.32 800 4.97 10.25 0.663 Mullite 1.66 2.69 0.32 400 3.18 7.62 1.424 Alumina 3.31 3.63 0.33 400 1.97 7.29 0.812 Alumina 3.31 3.63 0.33 800 4.81 10.61 0.316 Mullite 3.03 2.71 0.40 800 3.56 11.50 0.306 Mullite 3.45 2.74 0.41 800 2.98 10.52 0.306 Glass 3.11 2.48 0.41 800 6.82 10.23 0.302 Glass 3.11 2.48 0.50 800 8.71 10.18 0.266 Glass 3.11 2.48 0.51 800 29.45 9.48 0.227 Glass 3.11 2.48 0.53 800 19.51 10.10 0.257 Glass 3.11 2.48 0.54 800 8.87 10.80 0.272 Mullite 2.91 2.69 0.63 1500 14.72 7.55 0.306 Mullite 2.91 2.69 0.63 800 10.62 10.80 0.255 Glass 3.11 2.48 0.69 800 32.84 9.95 0.221 Glass 3.11 2.48 0.72 800 40.76 9.24 0.213 Glass 3.11 2.48 0.75 800 34.42 9.89 0.224 Glass 3.11 2.48 0.82 800 40.97 9.70 0.218

FIG. 47 shows a plot of media wear contamination of the product (% MOF) and product quality as measured by FLT tensile index, for all examples in Table 43, at a constant specific energy input of 3000 kWh/t. It is desirable to maximise FLT and minimise MOF, and therefore, the most desirable area is the top left side of the graph. Different data point symbols are used based upon the surface roughness of the grinding media.

As can be seen, for low roughnesses, there is a very wide spread of product qualities, which are due to the differing influence of media size and impeller speed between the trials. Ultimately, a good product can still be achieved, but it clearly requires a careful choice of the other operating conditions, since breaking the fibres in the absence of a high degree of roughness is more difficult, leading to many examples of very poor FLT values.

In contrast, grinding media of moderate roughnesses generally gives a higher FLT, and also the minimum FLT values in those series are considerably higher than those achieved at low roughnesses (i.e., using a finer media size than optimum has less severe consequences with rougher media than with smoother media, since the higher roughness partially compensates for the reduction in fibre breakage). This comes at a consequence of higher media wear rates, however.

Eventually, with further increasing surface roughness (see the Ra>0.4 μm series), the FLT quality achieved starts to decline compared to moderate media roughnesses. This, in consequence with extreme levels of media wear under such conditions, makes it clear that they are beyond the range of sensible choices of media roughness.

Since media wear is so intimately linked to the surface roughness, any improvements in tensile index achieved at a given energy input using media of relatively low surface roughness are particularly valuable, which is why the examples shown in Example 18 using smooth media, with FLT values of 10+Nm/g and % MOF values between about 0.5% and about 2% are particularly important. Primarily, these improvements were due to an appropriately selected media size and fibre solids content.

Table 44 below is a subset of Table 43, showing data for grinding media sizes in the range of 1.5-1.85 mm. Although this is an appropriate media size to use for many hardwoods, in the case of strong softwoods such as the NBSK used in this example, it is more difficult to efficiently obtain a good product. The lower roughness medias trialled in these cases were glass and ceria-zirconia. Table 44 shows that a range of impeller speeds were trialled, yet only at very high impeller speeds was MFC of even a moderate quality generated. This is evidently because the fine grinding media lacks the momentum necessary to break fibres and efficiently reduce Lc (I), unless a very high impeller speed is used.

However, in contrast to this, the use of rough mullite-based grinding media compensates for the fine media size, and so enables the generation of a high quality product at moderate impeller speeds (compare an impeller speed of 800 RPM, wherein the ceria-zirconia with a Ra of 0.07 μm gave an FLT of 4.74 Nm/g, to mullite media, with a Ra of 0.32 μm giving an FLT of 10.25 Nm/g). Using rougher grinding media extends the practical lower limit of media size downwards, although as seen in Tables 29 and 30, it comes at a cost of higher media wear content, which limits the applications in which the product can be used.

TABLE 44 Grinding media properties and impeller speed for lab-scale batch grinds performed with media sizes between 1.5 mm and 1.85 mm, and the resultant effect on media wear, MFC tensile index, and MFC fibre length Lc(l). MFC MFC length- Tensile weighted Media Media Media Impeller Media Index fibre Media Size Density Roughness Speed Wear FLT length Material (mm) (g/cm3) Ra (μm) (RPM) (% MOF) (Nm/g) Lc(l) (mm) Glass 1.81 2.48 0.028 800 0.76 3.54 1.66 Ceria-zirconia 1.51 6.04 0.069 300 0.89 3.47 1.80 Ceria-zirconia 1.51 6.04 0.069 500 1.48 3.30 1.72 Ceria-zirconia 1.51 6.04 0.069 800 0.22 4.74 1.40 Ceria-zirconia 1.51 6.04 0.069 1100 0.63 5.76 1.08 Ceria-zirconia 1.51 6.04 0.069 1500 0.63 7.64 0.50 Mullite 1.66 2.69 0.324 800 4.97 10.25 0.66 Mullite 1.66 2.69 0.324 400 3.18 7.62 1.42

Example 20. Batch Pilot-Scale Grinds Carried Out On NBSK Fibres, Comparing Media of Different Sizes and Roughnesses

TABLE 45 Pilot-scale batch grind parameters. Scale Pilot Scale Mode Batch Fibre NBSK (Metsa Botnia Pine) POP 100% Fibre Solids 0.5%-1.5% Intensity 50 kW/m3-74 kW/m3 MVC (%) 42-48 Energy Inputs (kWh/t) 500-8000 kWh/t

This example consists of various batch pilot-scale grinds carried out on NBSK fibres, comparing media of different sizes and roughnesses. These data are shown in Table 46 below. 5 mm grinding media with two different roughnesses are compared; in this case, there is no obvious advantage from using the rougher grinding media; the FLT tensile index of the product at a given energy input is no higher for the rougher media, yet the media wear content is much worse. Additionally, 5 mm, 4 mm, 3 mm, 1.7 mm and 1 mm media of largely similar roughnesses were compared. The FLT tensile index versus energy input and the media wear (% MOF) versus energy input are shown as FIGS. 45 and 46, respectively.

TABLE 46 A comparison of pilot-scale batch energy sweeps with media of various sizes and roughnesses for NBSK fibres ground in the absence of mineral. Note that where FLT data is missing, the data were not collected as it was considered that the grinder discharge solids were too low for the product to be a representative sample. MFC Specific Tensile Fibre Media Energy Index Media POP Solids Intensity MVC Media Media Roughness Input FLT Wear (%) (%) (kW/m3) (%) Composition Size Ra (μm) (kWh/t) (Nm/g) (% MOF) 100% 1.5% 65 48 Mullite 5 mm 0.30 500 3.40 1.12 750 4.64 1.55 1000 5.67 2.05 1250 7.17 2.80 1500 8.70 4.17 1750 9.12 5.77 2000 9.45 6.21 100% 1.00% 73 48 Mullite 5 mm 0.12 1000 6.78 0.74 2000 9.67 0.63 2150 11.26 0.73 3000 11.74 1.42 4000 11.66 1.69 5000 12.38 2.04 100% 1.00% 73 48 Mullite 4 mm 0.16 2000 9.05 1.38 3000 12.18 2.08 4000 12.29 2.76 5000 12.67 3.49 100% 1.00% 50 48 Mullite 3 mm 0.21 2500 10.55 2.64 3000 11.73 3.55 3500 12.27 3.28 4000 11.94 3.72 4500 11.51 4.26 5000 10.48 5.00 100% 1.00% 50 42 Mullite 1.7 mm 0.10 4000 11.13 0.89 4500 11.69 0.87 5000 13.24 1.11 5500 13.47 1.11 6000 13.56 1.08 100% 0.5% 50 42 Mullite 1 mm 0.082 1000 1.14 2000 1.48 3000 1.71 4000 1.82 5000 2.98 6000 3.31 7000 13.25 4.36 8000 13.94 3.77

In FIG. 48, the lowest energy input shown in each series is the energy above which the grinder was able to appropriately discharge the product. From this figure, it can be seen that the energy input at which the grinder could discharge is lower for larger media, and very high for finest media (i.e. 1 mm media). Additionally, the FLT achieved at this energy input peaks at somewhat higher values for finer media, but costs a higher energy input to obtain it.

Depending on what the target quality is for a certain application, there would be a different optimum media size. For example, if an FLT of 13 Nm/g is desirable, 1.7 mm media is the most efficient, achieving this at an energy input of 5000 kWh/t. However, if an FLT of 10 Nm/g is required, then 5 mm media is the most efficient, achieving this by 2150 kWh/t. The optimum FLT to select will depend on the FLT tensile index required for a given application.

Media wear versus roughness is shown in FIG. 46. As can be seen, the rougher media have a higher media wear rate than the smoother media, as determinable from the gradient of % MOF versus energy input. Considering a general requirement to reduce wear, media with low roughness can be utilized, but to do this for long, strong fibres such as NBSK, media size needs to be sufficiently large to efficiently fibrillate the fibres and permit discharge from the grinder. This is contrasted with the data in Table 44 for Example 19, which demonstrate with these fibres that fine media can only achieve a high quality product at reasonable energy input if rough media is used, with the consequence of greatly increased worn media content.

Example 21. Lab-Scale Grinds Carried Out on Eucalyptus and Acacia Fibres, Comparing Different MVCs at Two Media Sizes and Roughnesses

TABLE 47 Lab-scale batch grind parameters for Example 21. Parameter Set 1 Set 2 Scale Lab-scale Lab-scale Mode Batch Batch Fibre Eucalyptus Acacia POP 100% 100% Fibre Solids  1%  1% Impeller Speed Varied with MVC Varied with MVC to maintain to maintain 60 W power draw 120 W power draw MVC (%) 25-75 25-75 Energy Inputs (kWh/t) 2250 2250 Media Size (mm) 1.7 3 Media Roughness Ra (μm) 0.10 0.63

Two series of lab-scale grinds were carried out; one with eucalyptus fibres using 1.7 mm kaolin-based media with a surface roughness of about 0.1 μm, and another with acacia fibres using 3 mm kaolin-based media with a surface roughness of about 0.63 μm, in both cases with a specific energy input of 2250 kWh/t. The total volume in the grinder was maintained at constant, and the media volume concentration (MVC) was varied between experiments between 25%, 35%, 45%, 55%, 65%, and 75%. Impeller speed was varied with MVC to maintain a constant power draw between experiments in the series. After the target energy input was obtained, sufficient water was mixed into the vessel, to allow the media to be separated from the product by gravity. The MFC product was collected, and tested for FLT and measured with the fibre analyser.

FIG. 49 shows how the FLT varies with MVC applied. As can be seen, for eucalyptus fibres with fine, smooth media, under such conditions, the low MVC grinds (35% and especially 25%) give very low FLT index, whereas at 45% MVC and above, FLT is broadly similar. This is also the case for acacia fibres with coarse, rough media, though the differences between the different MVCs is less significant.

The effect on media wear rate is seen as FIG. 50. Although it appears that at higher MVC the product quality remains roughly constant, the acacia fibres with coarse rough media data demonstrates an increase in media wear at higher MVC values, thereby limiting the utility of these conditions. The eucalyptus fibres with fine smooth media data gave similarly very low wear across the MVC range, making the effect of MVC on wear on this case difficult to distinguish.

Although it may seem that higher MVCs are desirable, increasing MVC excessively makes discharge from the grinders in continuous mode increasingly difficult since there is insufficient free slurry to separate from the media. Thus, the upper limit of the preferred range of operation is expected to be in the region of 55%-65%.

Example 22. Lab-scale Grinds Carried Out on Acacia Fibres, Comparing Different Media Sizes and Surface Roughnesses

TABLE 48 Lab-scale batch grind parameters for Example 22. Scale Lab-scale Mode Batch Fibre Acacia POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 Media Size (mm) 0.25-10   Media Roughness Ra (μm) 0.02-0.63

Lab-scale batch grinds were carried out using acacia fibres using media of varying sizes, compositions, and surface roughnesses using the input parameters shown in Table 48. Table 49 shows the identities and key parameters of all media used in this study, along with key results in terms of FLT, fibre length, breakage factor (a computation of fibre length), and media wear particles in the product (as % MOF)

TABLE 49 Input conditions and key results of lab-scale grinds using acacia fibres using grinding media of different compositions, sizes, and roughnesses. Acacia feed fibres had an Lc(l) value of 0.766 mm, used as an input for the breakage factor calculations. Lc(l) Length- Specific FLT weighted Media Media Energy Tensile Media Average Breakage Media Size Roughness Input Index Wear Fibre Length Factor Composition (mm) Ra(μm) (kWh/t) (Nm/g) (MOF %) (mm) (1/mm) Kaolin-based 10 0.581 1500 1.4 N.M. 0.62 0.31 Glass 6 0.024 1500 2.9 N.M. 0.367 1.42 Kaolin-based 5 0.118 1500 3.7 N.M. 0.354 1.52 Kaolin-based 4 0.159 1500 3.9 N.M. 0.354 1.52 Kaolin-based 3 0.204 1500 4.9 N.M. 0.291 2.13 Glass 3 0.038 1500 5.1 N.M. 0.337 1.66 Kaolin-based 3 0.628 1500 3.9 N.M. 0.24 2.86 Kaolin-based 2.3 0.105 1500 5.6 N.M. 0.32 1.82 Kaolin-based 1.7 0.104 1500 7.1 N.M. 0.285 2.20 Kaolin-based 1.7 0.07 1500 7.0 N.M. 0.304 1.98 Glass 1.7 0.02 1500 6.6 N.M. 0.363 1.45 Kaolin-based 1 0.082 1500 8.3 N.M. 0.513 0.64 Glass 0.8 0.02 1500 6.1 N.M. 0.617 0.32 Zirconia-based 0.5 0.026 1500 4.7 N.M. 0.736 0.05 Zirconia-based 0.25 0.026 1500 3.1 N.M. 0.827 −0.10# Kaolin-based 10 0.581 3000 1.5 20.84 0.532 0.57 Glass 6 0.024 3000 4.7 3.87 0.247 2.74 Kaolin-based 5 0.118 3000 4.9 0.92 0.22 3.24 Kaolin-based 4 0.159 3000 5.3 1.23 0.24 2.86 Kaolin-based 3 0.204 3000 5.9 3.41 0.201 3.67 Glass 3 0.038 3000 6.9 3.04 0.227 3.10 Kaolin-based 3 0.628 3000 5.2 24.53 0.169 4.61 Kaolin-based 2.3 0.105 3000 7.4 0.69 0.225 3.14 Kaolin-based 1.7 0.104 3000 9.5 2.67 0.173 4.47 Kaolin-based 1.7 0.07 3000 9.1 1.35 0.185 4.10 Glass 1.7 0.02 3000 9.6 0.26 0.212 3.41 Kaolin-based 1 0.082 3000 11.1 0.93 0.263 2.50 Glass 0.8 0.02 3000 9.4 1.19 0.399 1.20 Zirconia-based 0.5 0.026 3000 7.0 N/A 0.607 0.34 Zirconia-based 0.25 0.026 3000 3.4 0.85 0.805 −0.06# For these very fine media sizes for glass and zirconia, it was not possible to measure roughness using the white light interferometer due to too great curvature at the media surface, so roughness values here are extrapolated from those of larger sizes with identical composition (these media are made by melting the particles with surface tension drawing them into spheres, so there is no reason to expect variance in roughness with sizes; in fact, there is little measured variation in sizes above the measurement limit). Media wear so low that it is below the practical measurement limit of the test. #Negative breakage factor values because product particles appear longer than feed particles. This is because the media is incompetent at breaking the larger fibres, with only the smaller particles being broken below the fibre analyser measurement limit. Media wear was measured only for the 3000 kWh/t samples.

FIGS. 51 and 52 show the breakage factor vs. media size for the 1500 kWh/t and 3000 kWh/t grinds respectively, with different data series depending on media composition and roughness. As can be seen, for smooth kaolin media (Ra=0.08 μm-0.2 μm) the breakage factor appears to peak at a media size of approx. 1.7 mm, and declines at sizes above this, whereas for very smooth glass media (Ra<0.4 μm), the peak is around 3 mm. Breakage factor rapidly accelerates between media sizes of 0.5 mm and 1.7 mm, though excessively fine media (0.25 mm) is incompetent at fibre breakage regardless of energy input.

The influence of surface roughness on fibre breakage can be seen both from FIGS. 51 and 52, and in Table 49, when comparing the data for media sizes where multiple roughnesses are investigated (1.7 mm and 3 mm media). As can be expected, breakage factor is enhanced with increasing roughnesses.

The effect of media size on FLT tensile index are shown as FIGS. 53 and 54 for the 1500 kWh/t and 3000 kWh/t data, respectively. Here, a clear peak is seen around a media size of 1 mm media, with sharp declines in FLT above and below this. It is obvious from this that media sizes below about 0.5 mm and above 3 mm give particularly poor product quality.

Additionally, despite media roughness increasing breakage factor considerably, rougher media does not correspond to an improvement in FLT, and in fact for the 3 mm media a rougher media results in a decline in quality. This can be rationalized as the media sizes tested (1.7 mm and especially 3 mm), fibre breakage factor is already very high with smooth media; any enhancements in breakage factor due to an increase in roughness is superfluous since clearly the fibrillation process is not being limited by the breakage of the fibres; instead, the more aggressive conditions result in excessive damage to the fibrils and a decline in product quality.

The peak in FLT occurs at lower media sizes than the peak in fibre breakage; this supports the discoveries demonstrated elsewhere that media size needs to be coarse enough to cause enough fibre breakage to advance the fibrillation process, but not excessively high that excess fibril damage is caused.

As demonstrated elsewhere in these examples, for a given media composition, media surface roughness correlates well with media wear content in the product (% MOF), as seen from the data in Table 49.

Considering both the goals of maximising FLT and minimising media wear at a given energy input, FIG. 55 shows a plot of % MOF vs. FLT, annotated with both media size (1st number in label) and surface roughness (2nd number in label). A high FLT and low % MOF is most desirable, so moving left and up on FIG. 55 represent improvements. Here, it is obvious that for acacia fibres, using relatively fine media (2.3 mm and below) improves FLT and using relatively smooth media lowers % MOF without noticeably degrading FLT.

From these data, this implies that acacia fibres, and other similarly weak, short hardwood fibres, could be optimally processed using media roughnesses of about 0.2 μm and below, and media sizes from about 0.5 mm to about 3 mm, with outside these ranges either being undesirable due to poor FLT and/or high levels of media wear.

Note that operating a multi-stage cascade process is expected to widen the operational window, by allowing coarser first stage media sizes to be used as the fibrillation is compensated for in a finer second stage, or by allowing finer second stages to be used, as the fibre breakage is compensated for in a coarser first stage.

Example 23. Lab-Scale Grinds Carried Out on Eucalyptus Fibres, Comparing Different Media Sizes and Surface Roughnesses

TABLE 50 Lab-scale batch grind parameters for Example 23. Scale Lab-scale Mode Batch Fibre Eucalyptus POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 Media Size (mm)  0.5-10 Media Roughness Ra (μm) 0.02-0.8

Lab-scale batch grinds were carried out using Eucalyptus fibres using media of varying sizes, compositions, and surface roughnesses using the input parameters shown in Table 50. Table 51 shows the identities and key parameters of all media used in this study, along with key results in terms of FLT, fibre length, breakage factor (a computation of fibre length), and media wear particles in the product (as % MOF).

TABLE 51 Input conditions and key results of lab-scale grinds using Eucalyptus fibres using grinding media of different compositions, sizes, and roughnesses. Eucalyptus feed fibres had an Lc(l) value of 0.721 mm, used as an input for the breakage factor calculations. Lc(l) Length- Specific FLT weighted Media Media Energy Tensile Media Average Breakage Media Size Roughness Input Index Wear Fibre Length Factor Composition (mm) Ra (μm) (kWh/t) (Nm/g) (MOF %) (mm) (1/mm) Kaolin-based 10 0.581 1500 1.9 N.M. 0.63 0.20 Glass 6 0.024 1500 5.0 N.M. 0.47 0.74 Kaolin-based 5 0.118 1500 7.0 N.M. 0.4 1.11 Kaolin-based 4 0.159 1500 7.9 N.M. 0.417 1.01 Kaolin-based 3 0.204 1500 8.4 N.M. 0.415 1.02 Kaolin-based 3 0.122 1500 8.4 N.M. 0.447 0.85 Glass 3 0.038 1500 8.0 N.M. 0.471 0.74 Kaolin-based 3 0.628 1500 8.8 N.M. 0.325 1.69 Kaolin-based 3 0.797 1500 6.9 N.M. 0.392 1.16 Kaolin-based 2.3 0.105 1500 8.5 N.M. 0.464 0.77 Kaolin-based 1.7 0.104 1500 8.4 N.M. 0.566 0.38 Glass 1.7 0.02 1500 8.9 N.M. 0.585 0.32 Kaolin-based 1 0.082 1500 8.8 N.M. 0.648 0.16 Glass 0.8 0.02 1500 6.0 N.M. 0.693 0.06 Zirconia-based 0.5 0.026 1500 5.9 N.M. 0.731 −0.02# Kaolin-based 10 0.581 3000 1.9 30.79 0.56 0.40 Glass 6 0.024 3000 7.0 3.42 0.365 1.35 Kaolin-based 5 0.118 3000 10.3 1.46 0.292 2.04 Kaolin-based 4 0.159 3000 9.8 1.57 0.322 1.72 Kaolin-based 3 0.204 3000 9.6 3.97 0.279 2.20 Kaolin-based 3 0.122 3000 10.6 2.72 0.316 1.78 Glass 3 0.038 3000 10.1 1.96 0.341 1.55 Kaolin-based 3 0.628 3000 11.0 21.60 0.229 2.98 Kaolin-based 3 0.797 3000 9.8 73.11 0.263 2.42 Kaolin-based 2.3 0.105 3000 11.0 0.80 0.341 1.55 Mullite 1.7 0.104 3000 12.4 1.07 0.392 1.16 Glass 1.7 0.02 3000 11.4 0.69 0.412 1.04 Kaolin-based 1 0.082 3000 12.6 1.51 0.513 0.56 Glass 0.8 0.02 3000 9.6 1.04 0.593 0.30 Zirconia-based 0.5 0.026 3000 8.5 0.93 0.656 0.14 For these very fine media sizes for glass and zirconia, it was not possible to measure roughness using the white light interferometer due to too great curvature at the media surface, so roughness values here are extrapolated from those of larger sizes with identical composition (these media are made by melting the particles with surface tension drawing them into spheres, so there is no reason to expect variance in roughness with sizes; in fact, there is little measured variation in sizes above the measurement limit). #Negative breakage factor values because product particles appear longer than feed particles. This is because the media is incompetent at breaking the larger fibres, with only the smaller particles being broken below the fibre analyser measurement limit. Media wear was measured only for the 3000 kWh/t samples.

FIGS. 56 and 57 show the breakage factor vs. media size for the 1500 kWh/t and 3000 kWh/t grinds respectively, with different data series depending on media composition and roughness. As can be seen, the breakage factor appears to peak at a media size of approx. 3-5 mm, and declines at sizes above this. Breakage factor rapidly declines at media sizes below approx. 3 mm, with media sizes below approx. 1 mm being particularly incompetent. It is also apparent that for media sizes where multiple media roughnesses are trialled (3 mm and 1.7 mm), the rougher media increases the breakage factor.

The effect of media size on FLT tensile index are shown as FIGS. 58 and 59 for the 1500 kWh/t and 3000 kWh/t data, respectively. At 1500 kWh/t, media sizes between 1 mm and 4 mm give similarly high FLT values, with declines in quality seen outside these ranges. At 3000 kWh/t, declines in quality at the coarser media sizes is more obvious, but it is still relatively minor compared to a weaker fibre like acacia.

As is seen in other examples, even when the fibre breakage factor is significantly below the peaks seen in FIGS. 56 and 57, an excellent product can still be made; in fact, the peak in FLT once again occurs at a lower media size than the peak in breakage factor. However, it is obvious that an excessively low breakage factor, i.e. using 0.5 mm zirconia media, excessively inhibits the fibrillation step and leads to relatively ineffective grinding. As demonstrated in other examples, excessively coarse media inhibits fibrillation, and so a decline in FLT at coarser media sizes is also seen.

When considering the varying roughnesses at the 3 mm and 1.7 mm media data points in FIGS. 58 and 59, a rougher media is not showing a clear benefit in FLT at either energy input. This can be rationalized by the fact that at these media sizes, the breakage factor even with the smoother media that is selected is already sufficient for efficient fibrillation, and so increasing this by increasing media roughness does not provide significant benefit. Although no such examples are shown here, considering the theory developed, it is expected that rough media at sizes where fibre breakage factor becomes limiting (i.e. approx. 0.5-1 mm) should exhibit a substantial benefit in FLT compared to a smoother media.

As demonstrated elsewhere in these examples, for a given media composition, media surface roughness correlates well with media wear content in the product (% MOF). For these data, this plot is shown as FIG. 60, where an increase in media wear from differences from increasing media roughness is obvious when media composition is controlled for.

Considering both the goals of maximising FLT and minimising media wear at a given energy input, FIG. 60 shows a plot of % MOF vs. FLT, annotated with both media size (1st number in label) and surface roughness (2nd number in label). A high FLT and low % MOF is most desirable, so moving left and up on FIG. 60 represent improvements. Here, it is obvious that for eucalyptus fibres, using relatively fine media or moderate sized media (5 mm and below) improves FLT, and using relatively smooth media lowers % MOF without noticeably degrading FLT. Despite similar fibre geometry, due to the greater resistance of eucalyptus fibres to breaking compared to acacia fibres, it is much more permissible to use coarser media in the case of eucalyptus since it is more difficult to cause excessive damage to the fibres.

From these data, this implies that eucalyptus fibres, and other similarly strong, short hardwood fibres, could be optimally processed using media roughnesses of about 0.3 μm and below, and media sizes from about 0.8 mm to about 6 mm, with outside these ranges either being undesirable due to poor FLT and/or high levels of media wear.

Example 24. Lab-Scale Grinds Carried Out on NBSK Fibres, Comparing Different Media Sizes and Surface Roughnesses

TABLE 52 Lab-scale batch grind parameters for Example 24. Scale Lab-scale Mode Batch Fibre NBSK POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 Media Size (mm)  1-10 Media Roughness Ra (μm) 0.02-0.8

Lab-scale batch grinds were carried out using NBSK fibres using media of varying sizes, compositions, and surface roughnesses using the input parameters shown in Table 52. Table 53 shows the identities and key parameters of all media used in this study, along with key results in terms of FLT, fibre length, breakage factor (a computation of fibre length), and media wear particles in the product (as % MOF).

TABLE 53 Input conditions and key results of lab-scale grinds using NBSK fibres using grinding media of different compositions, sizes, and roughnesses. NBSK feed fibres had an Lc(l) value of 1.789 mm, used as an input for the breakage factor calculations. Lc(l) Length- Specific FLT weighted Media Media Energy Tensile Media Average Breakage Media Size Roughness Input Index Wear Fibre Length Factor Composition (mm) Ra (μm) (kWh/t) (Nm/g) (MOF %) (mm) (1/mm) Kaolin-based 10 0.581 1500 2.4 N.M. 0.964 0.48 Glass 6 0.024 1500 4.3 N.M. 0.778 0.73 Kaolin-based 5 0.118 1500 6.8 N.M. 0.646 0.99 Kaolin-based 4 0.159 1500 7.3 N.M. 0.772 0.74 Kaolin-based 3 0.204 1500 7.9 N.M. 0.917 0.53 Kaolin-based 3 0.122 1500 8.0 N.M. 1.023 0.42 Glass 3 0.038 1500 5.9 N.M. 1.003 0.44 Kaolin-based 3 0.628 1500 9.9 N.M. 0.428 1.78 Kaolin-based 3 0.797 1500 8.1 N.M. 0.502 1.43 Kaolin-based 2.3 0.105 1500 6.8 N.M. 1.289 0.22 Kaolin-based 1.7 0.104 1500 5.9 N.M. 1.584 0.07 Glass 1.7 0.02 1500 4.1 N.M. 1.641 0.05 Kaolin-based 1.7 0.324 1500 8.7 N.M. 1.042 0.40 Kaolin-based 1 0.082 1500 4.2 N.M. 1.816 −0.01# Kaolin-based 10 0.581 3000 3.3 24.09 0.631 1.03 Glass 6 0.024 3000 8.1 3.28 0.466 1.59 Kaolin-based 5 0.118 3000 9.9 1.10 0.444 1.69 Kaolin-based 4 0.159 3000 10.4 1.29 0.453 1.65 Kaolin-based 3 0.204 3000 11.1 2.32 0.491 1.48 Kaolin-based 3 0.122 3000 10.5 2.04 0.529 1.33 Glass 3 0.038 3000 8.5 0.80 0.549 1.26 Kaolin-based 3 0.628 3000 12.5 20.87 0.263 3.24 Kaolin-based 3 0.797 3000 11.0 81.44 0.29 2.89 Kaolin-based 2.3 0.105 3000 10.1 0.33 0.653 0.97 Kaolin-based 1.7 0.104 3000 8.8 0.58 1.095 0.35 Glass 1.7 0.02 3000 7.6 0.60 1.243 0.25 Kaolin-based 1.7 0.324 3000 12.8 11.62 0.412 1.87 Kaolin-based 1 0.082 3000 6.0 1.49 1.788 0.00 #Negative breakage factor values because product particles appear longer than feed particles. This is because the media is incompetent at breaking the larger fibres, with only the smaller particles being broken below the fibre analyser measurement limit. Media wear was measured only for the 3000 kWh/t samples.

FIGS. 61 and 62 show the breakage factor vs. media size for the 1500 kWh/t and 3000 kWh/t grinds, respectively, with different data series depending on media composition and roughness. As can be seen, the breakage factor appears to increase with media size across the whole range of sizes being tested when relatively smooth kaolin-based grinding media is used, with 5 mm media giving the greatest breakage rate. There is a decline at higher media sizes; 10 mm rough kaolin media is notably worse, and 6 mm glass media is somewhat worse (the latter case likely partly due to the lower surface roughness).

For relatively smooth kaolin-based media, media sizes of approx. 1.7 mm and below are relatively poor at fibre breakage. However, as is apparent in FIGS. 61 and 62, a higher surface roughness greatly enhances breakage factor; in the case of finer media (1.7 mm), using the rougher kaolin-based media increases breakage rate severalfold.

The effect of media size on FLT tensile index is shown as FIGS. 63 and 64 for the 1500 kWh/t and 3000 kWh/t data, respectively. Given similar roughness levels, there appears to be a peak in FLT achieved about a media size of approx. 3 mm, with sizes between 2 mm and 5 mm giving similarly good FLT values.

As is clear from the absolute values of the breakage factors in comparison to other fibres tested in these examples under otherwise identical conditions, NBSK fibres are relatively difficult to break. Consequently, even with media sizes of 1.7 mm and 3 mm media, fibre breakage can still be expected to be limiting the efficiency of FLT development. With this in mind, enhancements in fibre breakage due to an increased surface roughness could be expected to improve FLT obtained at a given energy input.

FIGS. 63 and 64 demonstrate that this is indeed the case, with the smoothest glass media giving lower FLT values for a given media size, and the rougher kaolin-based media giving the highest FLT values for a given media size. In fact, the best results achieved from this perspective are rough 1.7 mm media at 3000 kWh/t, which represents an approx. 70% increase in FLT compared to the smoothest media of that size (glass media).

Even with the coarser 3 mm media, a clear benefit is also seen, which is notably absent for the hardwood fibres discussed in other examples; this is believed to be the aforementioned limitation in breakage factor being overcome for NBSK fibres, whereas it is not a limitation at those media sizes for the hardwood fibres.

It must be noted that despite media roughness giving a clear benefit in these cases, there are clearly limitations to the improvement that can be obtained by increasing media size and roughness; notably, the rough 10 mm kaolin-based media data point is the worst example out of all sizes, clearly due to excessive fibril damage and insufficient fibrillation.

Considering these data, consideration of media surface roughness in this example broadens the window of practical media sizes considerably. For roughnesses of approx. 0.2 μm and below, media sizes of approx. 2 mm to approx. 8 mm may be the most efficient, but for higher roughnesses beyond this, considering the 1.7 mm media results discussed here, it is expected that the lower limit of this size could be lowered considerably, perhaps even to approx. 1 mm.

However, despite the improvements in FLT seen at higher roughnesses, this must be considered against the consequent higher level of media wear. FIG. 65 shows the plot of media wear vs. FLT at 3000 kWh/t for NBSK fibres. It can be seen that with appropriate selection of media size, rougher media ultimately gives the highest FLT, but it comes at the cost of much higher media wear rate.

For many situations, it is not expected that the increased media wear rate from using very rough media would be worth the quality improvement due to the cost of grinding media consumed and resultant product contamination, and it would instead be more preferable to use a smoother and coarser grinding media for somewhat lower FLT but much lower wear content. However, for situations, for example, when grinding media is very cheap and wear in the product is not expected to be particularly detrimental, a more optimal configuration would be moderately rough, fine media (approx. 1.7 mm size, approx. 0.3 μm roughness).

Example 25. Lab-Scale Grinds Carried Out on OCC Fibres, Comparing Different Media Sizes and Surface Roughnesses

TABLE 54 Lab-scale batch grind parameters for Example 25. Scale Lab-scale Mode Batch Fibre OCC (Old Corrugated Containers) POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 Media Size (mm)  0.5-10 Media Roughness Ra (μm) 0.02-0.6

Lab-scale batch grinds were carried out using OCC fibres using media of varying sizes, compositions, and surface roughnesses using the input parameters shown in Table 54. Table 55 shows the identities and key parameters of all media used in this study, along with key results in terms of FLT, fibre length, breakage factor (a computation of fibre length), and media wear particles in the product (as % MOF).

TABLE 55 Input conditions and key results of lab-scale grinds using OCC fibres using grinding media of different compositions, sizes, and roughnesses. OCC feed fibres had an Lc(l) value of 0.889 mm, used as an input for the breakage factor calculations. Lc(l) Length- Specific FLT weighted Media Media Energy Tensile Media Average Breakage Media Size Roughness Input Index Wear Fibre Length Factor Composition (mm) Ra (μm) (kWh/t) (Nm/g) (MOF %) (mm) (1/mm) Kaolin-based 10 0.581 1500 2.5 14.81 0.627 0.47 Glass 6 0.024 1500 4.5 1.89 0.359 1.66 Kaolin-based 5 0.118 1500 5.4 0.87 0.331 1.90 Kaolin-based 3 0.204 1500 6.5 1.96 0.336 1.85 Kaolin-based 2.3 0.105 1500 6.6 0.38 0.36 1.65 Kaolin-based 1.7 0.104 1500 8.0 0.87 0.369 1.59 Kaolin-based 1.7 0.07 1500 6.9 0.89 0.442 1.14 Kaolin-based 1 0.082 1500 6.7 1.79 0.695 0.31 Glass 0.8 0.02 1500 4.8 0.92 0.783 0.15 Zirconia-based 0.5 0.026 1500 4.2 N/A 0.915 −0.03 Kaolin-based 10 0.581 3000 2.4 29.62 0.271 2.57 Glass 6 0.024 3000 6.0 3.78 0.27 2.58 Kaolin-based 5 0.118 3000 7.3 1.74 0.242 3.01 Kaolin-based 3 0.204 3000 8.5 3.91 0.232 3.19 Kaolin-based 2.3 0.105 3000 8.8 0.76 0.272 2.55 Kaolin-based 1.7 0.104 3000 9.5 1.74 0.213 3.57 Kaolin-based 1.7 0.07 3000 8.9 1.78 0.266 2.63 Kaolin-based 1 0.082 3000 9.1 3.59 0.567 0.64 Glass 0.8 0.02 3000 6.5 1.85 0.616 0.50 Zirconia-based 0.5 0.026 3000 5.6 N/A 0.896 −0.01 For these very fine media sizes for glass and zirconia, it was not possible to measure roughness using the white light interferometer due to too great curvature at the media surface, so roughness values here are extrapolated from those of larger sizes with identical composition (these media are made by melting the particles with surface tension drawing them into spheres, so there is no reason to expect variance in roughness with sizes; in fact, there is little measured variation in sizes above the measurement limit). Media wear so low that it is below the practical measurement limit of the test. # Negative breakage factor values because product particles appear longer than feed particles. This is because the media is incompetent at breaking the larger fibres, with only the smaller particles being broken below the fibre analyser measurement limit. Media wear was measured only for the 3000 kWh/t samples.

FIGS. 66 and 67 show the breakage factor vs. media size for the 1500 kWh/t and 3000 kWh/t grinds, respectively, with different data series depending on media composition and roughness. As can be seen, the breakage factor appears to be highest in the region of 1.7 mm to 6 mm.

FIGS. 68 and 69 show the FLT versus media size at 1500 kWh/t and 3000 kWh/t, respectively. Here, the best results are seen with media sizes between about 1 mm and about 4 mm, with this peak region being at somewhat finer media sizes than the optimum for fibre breakage factor, as has been seen in many of the other examples.

FIG. 70 shows a plot of media wear versus FLT, with the best results apparent for media roughnesses of between 0.07-0.2 μm, and media sizes of about 1 mm to about 3 mm. Note that in the case of OCC, media wear as measured from ash after furnacing must be measured differently: recycled fibres contain some level of mineral, so it must be ashed at 450° C. to burn away only fibre without decomposing any of the minerals present. This must have the ash content of the unprocessed OCC fibres subtracted from it to give the mineral content that is due to worn grinding media. Table 55 shows that these values are about what is expected when considering the other fibre species, NBSK, eucalyptus, and acacia, in other examples.

Example 26. Lab-Scale Grinds Carried Out on Acacia, Eucalyptus, NBSK, and OCC Fibres, Comparing Different Media Sizes at Similar Surface Roughnesses

Data from Examples 22, 23, 24, and 25, have been combined here in order to compare different fibre species. FIGS. 71 and 72 show the breakage factors and FLT using data selected for examples where relatively smooth kaolin-based media is used (Ra 0.08 μm-0.2 μm). FIGS. 73 and 74 also shows breakage factor and FLT, but only for very smooth media (glass and zirconia, Ra<0.04 μm).

By comparing the breakage factor of the fibre under otherwise similar conditions, an assessment of the resistance of the fibre to breaking can be made. From this, FIG. 71 shows that NBSK is the strongest fibre, and Eucalyptus is weaker, but still quite resistant to breaking. OCC is much weaker than eucalyptus, and acacia is clearly the weakest fibre, particularly obvious when assessing breakage using the finest media size on this graph.

It is apparent that the resistance of the fibre to breaking advises on the appropriate grinding media size to select for optimum FLT. FIG. 72 shows that for acacia, the weakest fibre by breakage factor, finer media towards 1 mm is strongly favoured, with rapid decay in quality at larger media sizes. This is true for OCC and eucalyptus fibres also, where approx. 1-1.7 mm media is optimum, although the decay in quality at larger media sizes is much less severe.

It can also be seen that the strongest fibre, NBSK, has an optimum media size region centred around 3 mm, with sharp decreases in quality for 1.7 mm media and finer.

From these considerations, it appears obvious that fibre breakage factors close to or below zero (i.e. <0.1 mm−1) will give a severely sub-quality product compared to other conditions, since the fibrillation process is being strongly inhibited by the inability of media collisions to break the fibres. More broadly, a breakage factor below approx. 0.4 mm−1 appears to imply a lower-than-optimum media size (although it must be noted that this is at a relatively high energy of 3000 kWh/t, and in fact at lower energy inputs breakage factors will be lower even with the optimum media size, only because of a lower residence time for breakage).

FIG. 73 shows the breakage factors using very smooth media. Here, the relative weakness of acacia fibres is very clear, with all the other fibres being in the same order as in FIG. 71. With these media, there are more examples which have breakage factors below 0.4 mm−1, and this exclusively captures all media sizes that FIG. 74 demonstrate are below the optimum for maximising FLT.

When considered together, it is clear that the optimum media size to use is strongly dependent on the fibre properties, with the media size around which the optimum region is centred apparently varying between about 0.8 mm and about 6 mm, particularly dependent on the resistance of the fibre to breaking. It is important to note also that media surface roughness, by enhancing fibre breakage, plays a key role, as evidenced in numerous other examples; in particular, it decreases the optimum media size, although at a consequence of increased media wear contamination.

With this in mind, due to considerations of expected variations in fibre properties and allowable media wear content, optimum media surface roughness can vary between 0.02 μm and 1 μm, and optimum media size can vary between 0.5 mm and 8 mm.

Example 27. Influence of Fibre Solids

TABLE 56 Lab-scale batch grind parameters for Example 27. Scale Lab-scale Mode Batch Fibre Eucalyptus POP 100% Fibre Solids 0.2%, 1%, 2.5%, 6% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 2500 Media Size (mm) 2.3 Media Roughness Ra (μm) 0.1

Laboratory-scale batch grinds were carried out using eucalyptus fibres under the conditions in Table 56. The fibre solids were varied between experiments from 0.2% to 6%, with the resultant FLTs shown as FIG. 75. Here, excessively low fibre solids (0.2%) leads to a poor quality FLT; this is believed to be due to an increased fraction of media collisions where there are no fibres present to break (and so the collision energy is wasted). At excessively high fibre solids content (6%), FLT is also relatively low, which is believed to be due to excessive energy lost due to the product viscosity. The optimum product quality is obtained at fibre solids content intermediate between these two values, with both the 1% and 2.5% fibre solids examples exhibiting high FLT values.

Considering the proposed mechanisms, it is expected that such an optimum region would shift somewhat based upon the other process parameters (e.g. fibre properties, media size, media roughness, energy input), although such extreme cases such as 0.2% and 6% fibre solids are expected to be universally poor.

Example 28. Comparative Example Versus Previous Practice

Previous practice is exemplified in EP2350387B1, with the production process in Examples 22, 23, 25, and 26 in EP2350387B1 repeated, with the characterisation methods in the present disclosure (FLT, media wear content, fibre analyser tests) applied to these samples.

These were then compared to examples from the present disclosure made with the same fibre species and energy inputs, where the process was adjusted as per the present disclosure to either (i) produce a product essentially free from mineral content, (ii) lower media wear content, or (iii) improve the level of fibrillation as measured by FLT.

Table 57 below compares the previous practice samples to the optimised conditions in this present disclosure.

TABLE 57 Parameters used in grinds comparing prior art to present work. EP2350387B1 Ex. Parameters 23, 24, 26, 27 Present Disclosure Fibre Species NBSK, Eucalyptus, NBSK, Eucalyptus, Acacia, Birch Acacia, Birch POP (fibre dry mass   5% 100%  fraction) Mineral used IC55 (55% <2 μm) GCC None Fibre Solids 1.75%  1% MVC 50% 48% Impeller Speed (RPM) 1000 800 Media Composition Carbolite 16/20 Calcined kaolin-based (calcined kaolin-based) Average Media Size (mm) 1 mm for all fibres 3 mm for NBSK; 1 mm for Eucalyptus, Acacia, and Birch Media Roughness Ra (μm) Not measured, but visibly 0.08 μm for 1 mm media; very rough (3 mm equivalent 0.2 μm for 3 mm media was measured to be 0.8 μm) Specific Energy Input 2500 and 5000 2500 and 5000 (kWh/t)

Due to the presence of an amount of mineral present in the prior art examples from EP2350387B1 (these are produced at 5 POP), it was necessary to modify the experimental techniques to allow for a direct comparison with the mineral-free samples in the present disclosure.

Regarding the media wear content as % MOF, to measure in a mineral-free sample the dry sample may be ashed in a furnace at 950° C. to determine the mineral content. However, in the prior art examples, produced at 5 POP, there is the presence of a significant amount of mineral which if this test were to be carried out in that manner would not be possible to distinguish from the worn media particles.

However, since the mineral used in this case is CaCO3, it is possible to identify the fraction of mineral that is worn grinding media from the loss of ignition:

Firstly, the ash content is measured on the dry sample by furnacing at 450° C. for at least 2 hours, where all the fibre burns away, leaving behind only the CaCO3 mineral and media wear particles.

Secondly, the sample is returned to the furnace at a temperature of 950° C. for at least 4 hours. Here, essentially all of the CaCO3 decomposes to release CO2. The difference in mass between the 450° C. and 950° C. measurements is due to this CO2 loss, and from this the CaCO3 content can be calculated by considering that the relative molecular mass of CO2 is 44% of that of CaCO3. Once the CaCO3 mass is known, the mass of the worn media can be found by subtracting the CaCO3 mass from the total ash mass at 450° C. The % MOF calculation is then done by dividing this calculated mass of worn media by the mass of fibre present.

The FLT test should be carried out at approx. 20 POP, as the mineral content strongly influences the strength of the sheet. For the present work mineral-free cases, then IC55 GCC mineral is added until 20 POP is reached. For the 5 POP prior art examples, this is a problem, as it is not practically possible to physically segregate the mineral particles from the MFC particles (to go from 5 POP to 20 POP requires that almost 80% of the mineral is removed.

Therefore, for the 5 POP samples, sufficient hydrochloric acid was added to dissolve the calcium carbonate. The acid was titrated until the pH became acidic, at which point the calcium carbonate was considered to be sufficiently dissolved. The titration was stopped around a pH of 4, since going to very low pH values (i.e. pH 0-1) risks damaging the fibrils through hydrolysis.

It was then necessary to drain away the vast majority of the solution, since the reaction would produce a large amount of soluble CaCl2). To do this, after dissolving out the vast majority of the calcium carbonate, the remaining sample was added to a beaker, and diluted with tap water up to a total volume of 10 L. This was then drained using the vacuum filter used in the FLT test for sheet making, but with a 6 μm nylon mesh fixed over the filter paper to prevent fine MFC particles from passing through the filter paper. Once drainage was complete, the MFC and remaining insoluble mineral particles formed a cake on the nylon mesh, with almost all of the 10 L volume (and therefore the dissolved CaCl2)) having passed through the filter.

The retained cake was resuspended in tap water, and was tested for solids content and % POP, along with Valmet FS5 fibre analyser measurements. Since the fibre content was much higher than 20% POP, these samples could be diluted with IC55 mineral to 20 POP for the FLT test in the same manner as the mineral-free samples using conditions from the present disclosure.

The FLT and % MOF values for both the prior art samples and the present work samples for all four fibre species tested are shown as FIG. 76 for an energy input of 2500 kWh/t, and FIG. 77 for an energy input of 5000 kWh/t.

Both figures demonstrate that for all fibre species, the conditions in the present work generate at least an approximately equivalent FLT product, but with much lower media wear content (between 1-2 orders of magnitude lower). This much lower media wear content is compounded with the fact that the present conditions do not produce a product that is 95% by dry mass mineral particles; for many paper and packaging applications, the presence of minerals are highly deleterious to performance as they add considerable weight whilst degrading strength properties. The reduction in media wear by using smoother grinding media also contributes to this advantage that a mineral-free sample has (i.e. an increased amount of applications where that MFC would be suitable).

In terms of FLT, at 2500 kWh/t, NBSK, birch, and eucalyptus give approximately similar values with both the prior art and the present work, which is believed to be due to the improvements in breakage efficiency from the use of rougher media and high mineral content in the prior art, albeit at the cost of extreme levels of media wear. For weak fibres like acacia, this level of breakage is clearly excessive, and results in the prior art giving an FLT result around half of that of the improved conditions in the present work, so for this fibre there is both a benefit in product quality and wear reduction from using the teachings in this disclosure. For the other three fibres tested, in this example, the benefit is the media wear reduction and the production of a mineral-free product formulation.

At 5000 kWh/t, the conclusions in terms of wear reduction from using the conditions in this present work are the same, with similarly large reductions in wear. However, in this case, for all fibres tested, a considerable improvement in FLT is also seen, whereas the prior art examples are at best no better than they were at 2500 kWh/t. This is rationalized by the lower media roughness and lack of mineral content of the present work being less damaging to the product fibrils, and so increases the product quality that is ultimately achievable.

Example 29. Lab-Scale Grinds Carried Out on Eucalyptus and NBSK Fibres, Using Steatite (Talc-Based) Media

TABLE 58 Lab-scale batch grind parameters for Example 29. Scale Lab-scale Mode Batch Fibre NBSK, Eucalyptus POP 100% Fibre Solids  1% Impeller Speed (RPM) 800 MVC (%) 48 Energy Inputs (kWh/t) 1500, 3000 Media Composition Steatite (talc-based) Media Size (mm) 3 Media Roughness Ra (μm) 0.77

A similar alternative to mullite (kaolin-based) media is steatite media, which is made from calcining talc. This present example exhibits grinds carried out using steatite media with NBSK and Eucalyptus. The media used in this example was relatively rough, at Ra=0.77 μm. Results in terms of FLT, % MOF, Lc (I), and fibre breakage rate are shown in Table 59. As can be seen the results are equivalent to what is observed using rough kaolin-based media, exhibited in Examples 23 and 24.

TABLE 59 Inputs and results for Example 29. Lc(l) Length- Specific FLT weighted Media Media Energy Tensile Media Average Breakage Fibre Media Size Roughness Input Index Wear Fibre Length Factor Species Composition (mm) Ra (μm) (kWh/t) (Nm/g) (MOF %) (mm) (1/mm) NBSK Talc-based 3 0.77 1500 10.3 8.1 0.446 1.68 NBSK Talc-based 3 0.77 3000 12.4 18.7 0.28 3.01 Eucalyptus Talc-based 3 0.77 1500 9.1 11.0 0.331 1.63 Eucalyptus Talc-based 3 0.77 3000 9.1 13.5 0.233 2.90

Claims

1. A method for preparing microfibrillated cellulose having at least one of increased tensile properties and reduced media wear, the method comprising:

providing a suspension comprising pulp fibres in a liquid medium, wherein the fibre solids content of the suspension is about 0.3 wt % to about 4 wt %;
microfibrillating the pulp fibres by grinding the suspension in the presence of a grinding medium in one or more wet stirred media mills, wherein: the grinding medium comprises about 0.5 mm to about 12 mm particles, the grinding medium has an arithmetic surface roughness from about 0.02 μm to about 2 μm, the media volume concentration of the grinding medium is about 30% to about 65% of the total volume of the charge of the one or more wet stirred media mills, the grinding is performed with an energy input of less than about 6,000 kWh/t, and grinding the suspension results in a worn media content of less than about 20% by dry mass; and
recovering the microfibrillated cellulose from the one or more wet stirred media mills.

2. The method according to claim 1, wherein the liquid medium is aqueous or a hydro alcoholic mixture.

3. (canceled)

4. The method according to claim 1, wherein the pulp fibres are hardwood pulp fibres.

5-12. (canceled)

13. The method according to claim 1, wherein the pulp fibres are softwood pulp fibres.

14-19. (canceled)

20. The method according to claim 1, wherein the pulp fibres are recycled pulp fibres.

21. The method according to claim 1, wherein the pulp fibres are non-wood cellulosic fibres.

22. (canceled)

23. The method according to claim 1, wherein the pulp fibres are a combination of two or more of hardwood pulp fibres, softwood pulp fibres, recycled pulp fibres, and non-wood cellulosic pulp fibres.

24. The method according to claim 1, wherein the fibre solids content of the suspension is about 0.3 wt. %, about 0.4 wt. %, about 0.5 wt. %, about 0.6 wt. %, about 0.7 wt. %, about 0.8 wt. %, about 0.9 wt. %, about 1 wt. %, about 1.1 wt. %, about 1.2 wt. %, about 1.3 wt. %, about 1.4 wt. %, about 1.5 wt. %, about 1.6 wt. %, about 1.7 wt. %, about 1.8 wt. %, about 1.9 wt. %, about 2 wt. %, about 2.1 wt. %, about 2.2 wt. %, about 2.3 wt. %, about 2.4 wt. %, about 2.5 wt. %, about 2.6 wt. %, about 2.7 wt. %, about 2.8 wt. %, about 2.9 wt. %, about 3 wt. %, about 3.1 wt. %, about 3.2 wt. %, about 3.3 wt. %, about 3.4 wt. %, about 3.5 wt. %, about 3.6 wt. %, about 3.7 wt. %, about 3.8 wt. %, about 3.9 wt. %, or about 4 wt. %.

25. (canceled)

26. The method according to claim 1, wherein the one or more wet stirred media mills comprises one or more vertical screened grinders, one or more horizontal milling apparatuses, or one or more stirred media detritors.

27-28. (canceled)

29. The method according to claim 1, wherein the one or more wet stirred media mills is two or more wet stirred media mills connected in series, wherein the two or more wet stirred media mills are selected from the group consisting of a screened grinder, a stirred media detritor, a horizontal milling apparatus, and combinations thereof.

30-34. (canceled)

35. The method according to claim 1, wherein the one or more wet stirred media mills is first and second wet stirred media mills, and the first wet stirred media mill includes grinding medium having a particle size greater than a particle size of grinding medium included in the second wet stirred media mill.

36. The method according to claim 1, wherein the one or more wet stirred media mills is first and second wet stirred media mills, and the first wet stirred media mill includes grinding medium having a particle size substantially similar to a particle size of grinding medium included in the second wet stirred media mill.

37. The method according to claim 1, wherein the one or more wet stirred media mills is two or more wet stirred media mills connected in series, and an output of a first of the two or more wet stirred media mills has a fibre breakage factor of 0.2 mm−1-3 mm−1 as calculated using data measured by a fibre analyzer.

38. (canceled)

39. The method according to claim 1, wherein the grinding medium comprises about 0.5 mm, about 0.6 mm, about 0.7 mm, about 0.8 mm, about 0.9 mm, about 1 mm, about 1.1 mm, about 1.2 mm, about 1.3 mm, about 1.4 mm, about 1.5 mm, about 1.6 mm, about 1.7 mm, about 1.8 mm, about 1.9 mm, about 2 mm, about 2.1 mm, about 2.2 mm, about 2.3 mm, about 2.4 mm, about 2.5 mm, about 2.6 mm, about 2.7 mm, about 2.8 mm, about 2.9 mm, about 3 mm, about 3.1 mm, about 3.2 mm, about 3.3 mm, about 3.4 mm, about 3.5 mm, about 3.6 mm, about 3.7 mm, about 3.8 mm, about 3.9 mm, about 4 mm, about 4.1 mm, about 4.2 mm, about 4.3 mm, about 4.4 mm, about 4.5 mm, about 4.6 mm, about 4.7 mm, about 4.8 mm, about 4.9 mm, about 5 mm, about 5.1 mm, about 5.2 mm, about 5.3 mm, about 5.4 mm, about 5.5 mm, about 5.6 mm, about 5.7 mm, about 5.8 mm, about 5.9 mm, about 6 mm, about 6.1 mm, about 6.2 mm, about 6.3 mm, about 6.4 mm, about 6.5 mm, about 6.6 mm, about 6.7 mm, about 6.8 mm, about 6.9 mm, about 7 mm, about 7.1 mm, about 7.2 mm, about 7.3 mm, about 7.4 mm, about 7.5 mm, about 7.6 mm, about 7.7 mm, about 7.8 mm, about 7.9 mm, about 8 mm, about 8.1 mm, about 8.2 mm, about 8.3 mm, about 8.4 mm, about 8.5 mm, about 8.6 mm, about 8.7 mm, about 8.8 mm, about 8.9 mm, about 9 mm, about 9.1 mm, about 9.2 mm, about 9.3 mm, about 9.4 mm, about 9.5 mm, about 9.6 mm, about 9.7 mm, about 9.8 mm, about 9.9 mm, about 10 mm, about 10.1 mm, about 10.2 mm, about 10.3 mm, about 10.4 mm, about 10.5 mm, about 10.6 mm, about 10.7 mm, about 10.8 mm, about 10.9 mm, about 11 mm, about 11.1 mm, about 11.2 mm, about 11.3 mm, about 11.4 mm, about 11.5 mm, about 11.6 mm, about 11.7 mm, about 11.8 mm, about 11.9 mm, or about 12 mm particles.

40-42. (canceled)

43. The method according to claim 1, wherein the grinding medium comprises mullite, alumina, silicate, zirconia, glass, steatite, or a combination thereof.

44-45. (canceled)

46. The method according to claim 1, wherein the grinding medium has an arithmetic surface roughness of about 0.02 μm, about 0.03 μm, about 0.04 μm, about 0.05 μm, about 0.06 μm, about 0.07 μm, about 0.08 μm, about 0.09 μm, about 0.1 μm, about 0.2 μm, about 0.3 μm, about 0.4 μm, about 0.5 μm, about 0.6 μm, about 0.7 μm, about 0.8 μm, about 0.9 μm, about 1 μm, about 1.1 μm, about 1.2 μm, about 1.3 μm, about 1.4 μm, about 1.5 μm, about 1.6 μm, about 1.7 μm, about 1.8 μm, about 1.9 μm, or about 2 μm.

47. (canceled)

48. The method according to claim 1, wherein the grinding medium has a density of about 2 g/cm3 to about 8 g/cm3.

49. The method according to claim 1, wherein the grinding is performed with an energy input of about 500 kWhH/t, about 600 kWhH/t, about 700 kWhH/t, about 800 kWhH/t, about 900 kWhH/t, about 1,000 kWhH/t, about 1,100 kWhH/t, about 1,200 kWhH/t, about 1,300 kWhH/t, about 1,400 kWhH/t, about 1,500 kWhH/t, about 1,600 kWhH/t, about 1,700 kWhH/t, about 1,800 kWhH/t, about 1,900 kWhH/t, about 2,000 k WhH/t, about 2,100 kWhH/t, about 2,200 k WhH/t, about 2,300 kWhH/t, about 2,400 kWhH/t, about 2,500 kWhH/t, about 2,600 kWhH/t, about 2,700 kWhH/t, about 2,800 kWhH/t, about 2,900 kWhH/t, about 3,000 kWhH/t, about 3,100 kWhH/t, about 3,200 kWhH/t, about 3,300 k WhH/t, about 3,400 kWhH/t, about 3,500 kWhH/t, about 3,600 kWhH/t, about 3,700 kWhH/t, about 3,800 kWhH/t, about 3,900 kWhH/t, about 4,000 kWhH/t, about 4,100 kWhH/t, about 4,200 kWhH/t, about 4,300 kWhH/t, about 4,400 kWhH/t, about 4,500 kWhH/t, about 4,600 k WhH/t, about 4,700 kWhH/t, about 4,800 k WhH/t, about 4,900 kWhH/t, about 5,000 kWhH/t, about 5,100 kWhH/t, about 5,200 k WhH/t, about 5,300 kWhH/t, about 5,400 kWhH/t, about 5,500 kWhH/t, about 5,600 kWhH/t, about 5,700 kWhH/t, about 5,800 kWhH/t, about 5,900 kWhH/t, or about 6,000 kWhH/t.

50. (canceled)

51. The method according to claim 1, wherein the media volume concentration of the grinding medium is about 30%, about 35%, about 40%, about 45%, about 50%, about 55%, about 60%, or about 65% of the total volume of the charge of the one or more wet stirred media mills.

52. (canceled)

53. The method according to claim 1, wherein the worn media content is about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 1%, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, about 10%, 11%, about 12%, about 13%, about 14%, about 15%, about 16%, about 17%, about 18%, about 19%, or about 20%.

54-57. (canceled)

Patent History
Publication number: 20240368835
Type: Application
Filed: Apr 29, 2024
Publication Date: Nov 7, 2024
Inventors: Lewis TAYLOR (Par), Chris BONDS (St. Austell), Paymaan TAHAMTAN (Par), Ben BULSON (St. Austell), David Robert SKUSE (Truro)
Application Number: 18/649,019
Classifications
International Classification: D21D 1/28 (20060101); D21H 11/12 (20060101); D21H 11/14 (20060101); D21H 11/18 (20060101);